Merge pull request #14441 from open-webui/dev

0.6.12

.github/ISSUE_TEMPLATE/bug_report.yaml

@@ -89,9 +89,20 @@ body:
           required: true
         - label: I have included the Docker container logs.
           required: true
-        - label: I have listed steps to reproduce the bug in detail.
+        - label: I have **provided every relevant configuration, setting, and environment variable used in my setup.**
+          required: true
+        - label: I have clearly **listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup** (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
+          required: true
+        - label: |
+            I have documented **step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation**. My steps:
+            - Start with the initial platform/version/OS and dependencies used,
+            - Specify exact install/launch/configure commands,
+            - List URLs visited, user input (incl. example values/emails/passwords if needed),
+            - Describe all options and toggles enabled or changed,
+            - Include any files or environmental changes,
+            - Identify the expected and actual result at each stage,
+            - Ensure any reasonably skilled user can follow and hit the same issue.
           required: true
-
   - type: textarea
     id: expected-behavior
     attributes:
@@ -112,15 +123,25 @@ body:
     id: reproduction-steps
     attributes:
       label: Steps to Reproduce
-      description: Providing clear, step-by-step instructions helps us reproduce and fix the issue faster. If we can't reproduce it, we can't fix it.
+      description: |
+        Please provide a **very detailed, step-by-step guide** to reproduce the issue. Your instructions should be so clear and precise that anyone can follow them without guesswork. Include every relevant detail—settings, configuration options, exact commands used, values entered, and any prerequisites or environment variables.  
+        **If full reproduction steps and all relevant settings are not provided, your issue may not be addressed.**
+
       placeholder: |
-        1. Go to '...'
-        2. Click on '...'
-        3. Scroll down to '...'
-        4. See the error message '...'
+        Example (include every detail):
+        1. Start with a clean Ubuntu 22.04 install.
+        2. Install Docker v24.0.5 and start the service.
+        3. Clone the Open WebUI repo (git clone ...).
+        4. Use the Docker Compose file without modifications.
+        5. Open browser Chrome 115.0 in incognito mode.
+        6. Go to http://localhost:8080 and log in with user "test@example.com".
+        7. Set the language to "English" and theme to "Dark".
+        8. Attempt to connect to Ollama at "http://localhost:11434".
+        9. Observe that the error message "Connection refused" appears at the top right.
+
+        Please list each step carefully and include all relevant configuration, settings, and options.
     validations:
       required: true
-
   - type: textarea
     id: logs-screenshots
     attributes:

.github/workflows/release-pypi.yml

@@ -17,6 +17,10 @@ jobs:
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+      - name: Install Git
+        run: sudo apt-get update && sudo apt-get install -y git
       - uses: actions/setup-node@v4
         with:
           node-version: 22

CHANGELOG.md

@@ -5,6 +5,119 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.6.12] - 2025-05-29
+
+### Added
+
+- 🧩 **Custom Advanced Model Parameters**: You can now add your own tailor-made advanced parameters to any model, empowering you to fine-tune behavior and unlock greater flexibility beyond just the built-in options—accelerate your experimentation.
+- 🪧 **Datalab Marker API Content Extraction Support**: Seamlessly extract content from files and documents using the Datalab Marker API directly in your workflows, enabling more robust structured data extraction for RAG and document processing with just a simple engine switch in the UI.
+- ⚡ **Parallelized Base Model Fetching**: Experience noticeably faster startup and model refresh times—base model data now loads in parallel, drastically shortening delays in busy or large-scale deployments.
+- 🧠 **Efficient Function Loading and Caching**: Functions are now only reloaded if their content changes, preventing unnecessary duplicate loads, saving bandwidth, and boosting performance.
+- 🌍 **Localization & Translation Enhancements**: Improved and expanded Simplified, Traditional Chinese, and Russian translations, providing smoother, more accurate, and context-aware experiences for global users.
+
+### Fixed
+
+- 💬 **Stable Message Input Box**: Fixed an issue where the message input box would shift unexpectedly (especially on mobile or with screen reader support), ensuring a smooth and reliable typing experience for every user.
+- 🔊 **Reliable Read Aloud (Text-to-Speech)**: Read aloud now works seamlessly across messages, so users depending on TTS for accessibility or multitasking will experience uninterrupted and clear voice playback.
+- 🖼 **Image Preview and Download Restored**: Fixed problems with image preview and downloads, ensuring frictionless creation, previewing, and downloading of images in your chats—no more interruptions in creative or documentation workflows.
+- 📱 **Improved Mobile Styling for Workspace Capabilities**: Capabilities management is now readable and easy-to-use even on mobile devices, empowering admins and users to manage access quickly on the go.
+- 🔁 **/api/v1/retrieval/query/collection Endpoint Reliability**: Queries to retrieval collections now return the expected results, bolstering the reliability of your knowledge workflows and citation-ready responses.
+
+### Removed
+
+- 🧹 **Duplicate CSS Elements**: Streamlined the UI by removing redundant CSS, reducing clutter and improving load times for a smoother visual experience.
+
+## [0.6.11] - 2025-05-27
+
+### Added
+
+- 🟢 **Ollama Model Status Indicator in Model Selector**: Instantly see which Ollama models are currently loaded with a clear indicator in the model selector, helping you stay organized and optimize local model usage.
+- 🗑️ **Unload Ollama Model Directly from Model Selector**: Easily release memory and resources by unloading any loaded Ollama model right in the model selector—streamline hardware management without switching pages.
+- 🗣️ **User-Configurable Speech-to-Text Language Setting**: Improve transcription accuracy by letting individual users explicitly set their preferred STT language in their settings—ideal for multilingual teams and clear audio capture.
+- ⚡ **Granular Audio Playback Speed Control**: Instead of just presets, you can now choose granular audio speed using a numeric input, giving you complete control over playback pace in transcriptions and media reviews.
+- 📦 **GZip, Brotli, ZStd Compression Middleware**: Enjoy significantly faster page loads and reduced bandwidth usage with new server-side compression—giving users a snappier, more efficient experience.
+- 🏷️ **Configurable Weight for BM25 in Hybrid Search**: Fine-tune search relevance by adjusting the weight for BM25 inside hybrid search from the UI, letting you tailor knowledge search results to your workflow.
+- 🧪 **Bypass File Creation with CTRL + SHIFT + V**: When “Paste Large Text as File” is enabled, use CTRL + SHIFT + V to skip the file creation dialog and instantly upload text as a file—perfect for rapid document prep.
+- 🌐 **Bypass Web Loader in Web Search**: Choose to bypass web content loading and use snippets directly in web search for faster, more reliable results when page loads are slow or blocked.
+- 🚀 **Environment Variable: WEBUI_AUTH_TRUSTED_GROUPS_HEADER**: Now sync and manage user groups directly via trusted HTTP header, unlocking smoother single sign-on and identity integrations for organizations.
+- 🏢 **Workspace Models Visibility Controls**: You can now hide workspace-level models from both the model selector and shared environments—keep your team focused and reduce clutter from rarely-used endpoints.
+- 🛡️ **Copy Model Link**: You can now copy a direct link to any model—including those hidden from the selector—making sharing and onboarding others more seamless.
+- 🔗 **Load Function Directly from URL**: Simplify custom function management—just paste any GitHub function URL into Open WebUI and import new functions in seconds.
+- ⚙️ **Custom Name/Description for External Tool Servers**: Personalize and clarify external tool servers by assigning custom names and descriptions, making it easier to manage integrations in large-scale workspaces.
+- 🌍 **Custom OpenAPI JSON URL Support for Tool Servers**: Supports specifying any custom OpenAPI JSON URL, unlocking more flexible integration with any backend for tool calls.
+- 📊 **Source Field Now Displays in Non-Streaming Responses with Attachments**: When files or knowledge are attached, the "source" field now appears for all responses, even in non-streaming mode—enabling improved citation workflow.
+- 🎛 **Pinned Chats**: Reduced payload size on pinned chat requests—leading to faster load times and less data usage, especially on busy warehouses.
+- 🛠 **Import/Export Default Prompt Suggestions**: Enjoy one-click import/export of prompt suggestions, making it much easier to share, reuse, and manage best practices across teams or deployments.
+- 🍰 **Banners Now Sortable from Admin Settings**: Quickly re-order or prioritize banners, letting you highlight the most critical info for your team.
+- 🛠 **Advanced Chat Parameters—Clearer Ollama Support Labels**: Parameters and advanced settings now explicitly indicate if they are Ollama-specific, reducing confusion and improving setup accuracy.
+- 🤏 **Scroll Bar Thumb Improved for Better Visibility**: Enhanced scrollbar styling makes navigation more accessible and visually intuitive.
+- 🗄️ **Modal Redesign for Archived and User Chat Listings**: Clean, modern modal interface for browsing archived and user-specific chats makes locating conversations faster and more pleasant.
+- 📝 **Add/Edit Memory Modal UX**: Memory modals are now larger and have resizable input fields, supporting easier editing of long or complex memory content.
+- 🏆 **Translation & Localization Enhancements**: Major upgrades to Chinese (Simplified & Traditional), Korean, Russian, German, Danish, Finnish—not just fixing typos, but consistency, tone, and terminology for a more natural native-language experience.
+- ⚡ **General Backend Stability & Security Enhancements**: Various backend refinements ensure a more resilient, reliable, and secure platform for smoother operation and peace of mind.
+
+### Fixed
+
+- 🖼️ **Image Generation with Allowed File Extensions Now Works Reliably**: Ensure seamless image generation even when strict file extension rules are set—no more blocked creative workflows due to technical hiccups.
+- 🗂 **Remove Leading Dot for File Extension Check**: Fixed an issue where file validation failed because of a leading dot, making file uploads and knowledge management more robust.
+- 🏷️ **Correct Local/External Model Classification**: The platform now accurately distinguishes between local and external models—preventing local models from showing up as external (and vice versa)—ensuring seamless setup, clarity, and management of your AI model endpoints.
+- 📄 **External Document Loader Now Functions as Intended**: External document loaders are reliably invoked, ensuring smoother knowledge ingestion from external sources—expanding your RAG and knowledge workflows.
+- 🎯 **Correct Handling of Toggle Filters**: Toggle filters are now robustly managed, preventing accidental auto-activation and ensuring user preferences are always respected.
+- 🗃 **S3 Tagging Character Restrictions Fixed**: Tags for files in S3 now automatically meet Amazon’s allowed character set, avoiding upload errors and ensuring cross-cloud compatibility.
+- 🛡️ **Authentication Now Uses Password Hash When Duplicate Emails Exist**: Ensures account security and prevents access issues if duplicate emails are present in your system.
+
+### Changed
+
+- 🧩 **Admin Settings: OAuth Redirects Now Use WEBUI_URL**: The OAuth redirect URL is now based on the explicitly set WEBUI_URL, ensuring single sign-on and identity provider integrations always send users to the correct frontend.
+
+### Removed
+
+- 💡 **Duplicate/Typo Component Removals**: Obsolete components have been cleaned up, reducing confusion and improving overall code quality for the team.
+- 🚫 **Streaming Upsert in Pinecone Removed**: Removed streaming upsert references for better compatibility and future-proofing with latest Pinecone SDK updates.
+
+## [0.6.10] - 2025-05-19
+
+### Added
+
+- 🧩 **Experimental Azure OpenAI Support**: Instantly connect to Azure OpenAI endpoints by simply pasting your Azure OpenAI URL into the model connections—bringing flexible, enterprise-grade AI integration directly to your workflow.
+- 💧 **Watermark AI Responses**: Easily add a visible watermark to AI-generated responses for clear content provenance and compliance with EU AI regulations—perfect for regulated environments and transparent communication.
+- 🔍 **Enhanced Search Experience with Dedicated Modal**: Enjoy a modern, powerful search UI in a dedicated modal (open with Ctrl/Cmd + K) accessible from anywhere—quickly find chats, models, or content and boost your productivity.
+- 🔲 **"Toggle" Filter Type for Chat Input**: Add interactive toggle filters (e.g. Web Search, Image, Code Interpreter) right into the chat input—giving you one-click control to activate features and simplifying the chat configuration process.
+- 🧰 **Granular Model Capabilities Editor**: Define detailed capabilities and feature access for each AI model directly from the model editor—enabling tailored behavior, modular control, and a more customized AI environment for every team or use case.
+- 🌐 **Flexible Local and External Connection Support**: Now you can designate any AI connection—whether OpenAI, Ollama, or others—as local or external, enabling seamless setup for on-premise, self-hosted, or cloud configurations and giving you maximum control and flexibility.
+- 🗂️ **Allowed File Extensions for RAG**: Gain full control over your Retrieval-Augmented Generation (RAG) workflows by specifying exactly which file extensions are permitted for upload, improving security and relevance of indexed documents.
+- 🔊 **Enhanced Audio Transcription Logic**: Experience smoother, more reliable audio transcription—very long audio files are now automatically split and processed in segments, preventing errors and ensuring even challenging files are transcribed seamlessly, all part of a broader stability upgrade for robust media workflows.
+- 🦾 **External Document Loader Support**: Enhance knowledge base building by integrating documents using external loaders from a wide range of data sources, expanding what your AI can read and process.
+- 📝 **Preview Button for Code Artifacts**: Instantly jump from an HTML code block to its associated artifacts page with the click of a new preview button—speeding up review and streamlining analysis.
+- 🦻 **Screen Reader Support for Response Messages**: All chat responses are now fully compatible with screen readers, making the platform more inclusive and accessible for everyone.
+- 🧑‍💼 **Customizable Pending User Overlay**: You can now tailor the overlay title and content shown to pending users, ensuring onboarding messaging is perfectly aligned with your organization’s tone and requirements.
+- 🔐 **Option to Disable LDAP Certificate Validation**: You can now disable LDAP certificate validation for maximum flexibility in diverse IT environments—making integrations and troubleshooting much easier.
+- 🎯 **Workspace Search by Username or Email**: Easily search across workspace pages using any username or email address, streamlining user and resource management.
+- 🎨 **High Contrast & Dark Mode Enhancements**: Further improved placeholder, input, suggestion, toast, and model selector contrasts—including a dedicated placeholder dark mode—for more comfortable viewing in all lighting conditions.
+- 🛡️ **Refined Security for Pipelines & Model Uploads**: Strengthened safeguards against path traversal vulnerabilities during uploads—ensuring your platform’s document and model management remains secure.
+- 🌏 **Major Localization Upgrades**: Comprehensive translation updates and improvements across Korean, Bulgarian, Catalan, Japanese, Italian, Traditional Chinese, and Spanish—including more accurate AI terminology for clarity; your experience is now more natural, inclusive, and professional everywhere.
+- 🦾 **General Backend Stability & Security**: Multiple backend improvements (including file upload, command navigation, and logging refactorings) deliver increased resilience, better error handling, and a more robust platform for all users.
+
+### Fixed
+
+- ✅ **Evaluation Feedback Endpoint Reliability**: Addressed issues with feedback submission endpoints to ensure seamless user feedback collection on model responses.
+- 🫰 **Model List State Fixes**: Resolved issues where model status toggles in the workspace page might inadvertently switch or confuse state, making the management of active/inactive models more dependable.
+- ✍️ **Admin Signup Logic**: Admin self-signup experience and validation flow is smoother and more robust.
+- 🔁 **Signout Redirect Flow Improved**: Logging out now redirects more reliably, reducing confusion and making session management seamless.
+
+## [0.6.9] - 2025-05-10
+
+### Added
+
+- 📝 **Edit Attached Images/Files in Messages**: You can now easily edit your sent messages by removing attached files—streamlining document management, correcting mistakes on the fly, and keeping your chats clutter-free.
+- 🚨 **Clear Alerts for Private Task Models**: When interacting with private task models, the UI now clearly alerts you—making it easier to understand resource availability and access, reducing confusion during workflow setup.
+
+### Fixed
+
+- 🛡️ **Confirm Dialog Focus Trap Reliability**: The focus now stays correctly within confirmation dialogs, ensuring keyboard navigation and accessibility is seamless and preventing accidental operations—especially helpful during critical or rapid workflows.
+- 💬 **Temporary Chat Admin Controls & Session Cleanliness**: Admins are now able to properly enable temporary chat mode without errors, and previous session prompts or tool selections no longer carry over—delivering a fresh, predictable, and consistent temporary chat experience every time.
+- 🤖 **External Reranker Integration Functionality Restored**: External reranker integrations now work correctly, allowing you to fully leverage advanced ranking services for sharper, more relevant search results in your RAG and knowledge base workflows.
+
 ## [0.6.8] - 2025-05-10
 
 ### Added

Dockerfile

@@ -26,6 +26,9 @@ ARG BUILD_HASH
 
 WORKDIR /app
 
+# to store git revision in build
+RUN apk add --no-cache git
+
 COPY package.json package-lock.json ./
 RUN npm ci
 

README.md

@@ -73,7 +73,17 @@ Want to learn more about Open WebUI's features? Check out our [Open WebUI docume
       </a>
     </td>
     <td>
-      Does your interface have a backend yet?<br>Try <a href="https://n8n.io/">n8n</a>
+      N8N • Does your interface have a backend yet?<br>Try <a href="https://n8n.io/">n8n</a>
+    </td>
+  </tr>
+  <tr>
+    <td>
+      <a href="https://warp.dev/open-webui" target="_blank">
+        <img src="https://docs.openwebui.com/sponsors/logos/warp.png" alt="n8n" style="width: 8rem; height: 8rem; border-radius: .75rem;" />
+      </a>
+    </td>
+    <td>
+      Wrap • The intelligent terminal for developers
     </td>
   </tr>
 </table>

backend/open_webui/config.py

@@ -989,6 +989,26 @@ DEFAULT_USER_ROLE = PersistentConfig(
     os.getenv("DEFAULT_USER_ROLE", "pending"),
 )
 
+PENDING_USER_OVERLAY_TITLE = PersistentConfig(
+    "PENDING_USER_OVERLAY_TITLE",
+    "ui.pending_user_overlay_title",
+    os.environ.get("PENDING_USER_OVERLAY_TITLE", ""),
+)
+
+PENDING_USER_OVERLAY_CONTENT = PersistentConfig(
+    "PENDING_USER_OVERLAY_CONTENT",
+    "ui.pending_user_overlay_content",
+    os.environ.get("PENDING_USER_OVERLAY_CONTENT", ""),
+)
+
+
+RESPONSE_WATERMARK = PersistentConfig(
+    "RESPONSE_WATERMARK",
+    "ui.watermark",
+    os.environ.get("RESPONSE_WATERMARK", ""),
+)
+
+
 USER_PERMISSIONS_WORKSPACE_MODELS_ACCESS = (
     os.environ.get("USER_PERMISSIONS_WORKSPACE_MODELS_ACCESS", "False").lower()
     == "true"
@@ -1731,6 +1751,9 @@ QDRANT_API_KEY = os.environ.get("QDRANT_API_KEY", None)
 QDRANT_ON_DISK = os.environ.get("QDRANT_ON_DISK", "false").lower() == "true"
 QDRANT_PREFER_GRPC = os.environ.get("QDRANT_PREFER_GRPC", "False").lower() == "true"
 QDRANT_GRPC_PORT = int(os.environ.get("QDRANT_GRPC_PORT", "6334"))
+ENABLE_QDRANT_MULTITENANCY_MODE = (
+    os.environ.get("ENABLE_QDRANT_MULTITENANCY_MODE", "false").lower() == "true"
+)
 
 # OpenSearch
 OPENSEARCH_URI = os.environ.get("OPENSEARCH_URI", "https://localhost:9200")
@@ -1825,6 +1848,73 @@ CONTENT_EXTRACTION_ENGINE = PersistentConfig(
     os.environ.get("CONTENT_EXTRACTION_ENGINE", "").lower(),
 )
 
+DATALAB_MARKER_API_KEY = PersistentConfig(
+    "DATALAB_MARKER_API_KEY",
+    "rag.datalab_marker_api_key",
+    os.environ.get("DATALAB_MARKER_API_KEY", ""),
+)
+
+DATALAB_MARKER_LANGS = PersistentConfig(
+    "DATALAB_MARKER_LANGS",
+    "rag.datalab_marker_langs",
+    os.environ.get("DATALAB_MARKER_LANGS", ""),
+)
+
+DATALAB_MARKER_USE_LLM = PersistentConfig(
+    "DATALAB_MARKER_USE_LLM",
+    "rag.DATALAB_MARKER_USE_LLM",
+    os.environ.get("DATALAB_MARKER_USE_LLM", "false").lower() == "true",
+)
+
+DATALAB_MARKER_SKIP_CACHE = PersistentConfig(
+    "DATALAB_MARKER_SKIP_CACHE",
+    "rag.datalab_marker_skip_cache",
+    os.environ.get("DATALAB_MARKER_SKIP_CACHE", "false").lower() == "true",
+)
+
+DATALAB_MARKER_FORCE_OCR = PersistentConfig(
+    "DATALAB_MARKER_FORCE_OCR",
+    "rag.datalab_marker_force_ocr",
+    os.environ.get("DATALAB_MARKER_FORCE_OCR", "false").lower() == "true",
+)
+
+DATALAB_MARKER_PAGINATE = PersistentConfig(
+    "DATALAB_MARKER_PAGINATE",
+    "rag.datalab_marker_paginate",
+    os.environ.get("DATALAB_MARKER_PAGINATE", "false").lower() == "true",
+)
+
+DATALAB_MARKER_STRIP_EXISTING_OCR = PersistentConfig(
+    "DATALAB_MARKER_STRIP_EXISTING_OCR",
+    "rag.datalab_marker_strip_existing_ocr",
+    os.environ.get("DATALAB_MARKER_STRIP_EXISTING_OCR", "false").lower() == "true",
+)
+
+DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION = PersistentConfig(
+    "DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION",
+    "rag.datalab_marker_disable_image_extraction",
+    os.environ.get("DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION", "false").lower()
+    == "true",
+)
+
+DATALAB_MARKER_OUTPUT_FORMAT = PersistentConfig(
+    "DATALAB_MARKER_OUTPUT_FORMAT",
+    "rag.datalab_marker_output_format",
+    os.environ.get("DATALAB_MARKER_OUTPUT_FORMAT", "markdown"),
+)
+
+EXTERNAL_DOCUMENT_LOADER_URL = PersistentConfig(
+    "EXTERNAL_DOCUMENT_LOADER_URL",
+    "rag.external_document_loader_url",
+    os.environ.get("EXTERNAL_DOCUMENT_LOADER_URL", ""),
+)
+
+EXTERNAL_DOCUMENT_LOADER_API_KEY = PersistentConfig(
+    "EXTERNAL_DOCUMENT_LOADER_API_KEY",
+    "rag.external_document_loader_api_key",
+    os.environ.get("EXTERNAL_DOCUMENT_LOADER_API_KEY", ""),
+)
+
 TIKA_SERVER_URL = PersistentConfig(
     "TIKA_SERVER_URL",
     "rag.tika_server_url",
@@ -1849,6 +1939,12 @@ DOCLING_OCR_LANG = PersistentConfig(
     os.getenv("DOCLING_OCR_LANG", "eng,fra,deu,spa"),
 )
 
+DOCLING_DO_PICTURE_DESCRIPTION = PersistentConfig(
+    "DOCLING_DO_PICTURE_DESCRIPTION",
+    "rag.docling_do_picture_description",
+    os.getenv("DOCLING_DO_PICTURE_DESCRIPTION", "False").lower() == "true",
+)
+
 DOCUMENT_INTELLIGENCE_ENDPOINT = PersistentConfig(
     "DOCUMENT_INTELLIGENCE_ENDPOINT",
     "rag.document_intelligence_endpoint",
@@ -1887,6 +1983,11 @@ RAG_RELEVANCE_THRESHOLD = PersistentConfig(
     "rag.relevance_threshold",
     float(os.environ.get("RAG_RELEVANCE_THRESHOLD", "0.0")),
 )
+RAG_HYBRID_BM25_WEIGHT = PersistentConfig(
+    "RAG_HYBRID_BM25_WEIGHT",
+    "rag.hybrid_bm25_weight",
+    float(os.environ.get("RAG_HYBRID_BM25_WEIGHT", "0.5")),
+)
 
 ENABLE_RAG_HYBRID_SEARCH = PersistentConfig(
     "ENABLE_RAG_HYBRID_SEARCH",
@@ -1920,6 +2021,16 @@ RAG_FILE_MAX_SIZE = PersistentConfig(
     ),
 )
 
+RAG_ALLOWED_FILE_EXTENSIONS = PersistentConfig(
+    "RAG_ALLOWED_FILE_EXTENSIONS",
+    "rag.file.allowed_extensions",
+    [
+        ext.strip()
+        for ext in os.environ.get("RAG_ALLOWED_FILE_EXTENSIONS", "").split(",")
+        if ext.strip()
+    ],
+)
+
 RAG_EMBEDDING_ENGINE = PersistentConfig(
     "RAG_EMBEDDING_ENGINE",
     "rag.embedding_engine",
@@ -2126,6 +2237,12 @@ BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL = PersistentConfig(
 )
 
 
+BYPASS_WEB_SEARCH_WEB_LOADER = PersistentConfig(
+    "BYPASS_WEB_SEARCH_WEB_LOADER",
+    "rag.web.search.bypass_web_loader",
+    os.getenv("BYPASS_WEB_SEARCH_WEB_LOADER", "False").lower() == "true",
+)
+
 WEB_SEARCH_RESULT_COUNT = PersistentConfig(
     "WEB_SEARCH_RESULT_COUNT",
     "rag.web.search.result_count",
@@ -2151,6 +2268,7 @@ WEB_SEARCH_CONCURRENT_REQUESTS = PersistentConfig(
     int(os.getenv("WEB_SEARCH_CONCURRENT_REQUESTS", "10")),
 )
 
+
 WEB_LOADER_ENGINE = PersistentConfig(
     "WEB_LOADER_ENGINE",
     "rag.web.loader.engine",
@@ -2839,6 +2957,12 @@ LDAP_CA_CERT_FILE = PersistentConfig(
     os.environ.get("LDAP_CA_CERT_FILE", ""),
 )
 
+LDAP_VALIDATE_CERT = PersistentConfig(
+    "LDAP_VALIDATE_CERT",
+    "ldap.server.validate_cert",
+    os.environ.get("LDAP_VALIDATE_CERT", "True").lower() == "true",
+)
+
 LDAP_CIPHERS = PersistentConfig(
     "LDAP_CIPHERS", "ldap.server.ciphers", os.environ.get("LDAP_CIPHERS", "ALL")
 )

backend/open_webui/env.py

@@ -349,6 +349,10 @@ WEBUI_AUTH_TRUSTED_EMAIL_HEADER = os.environ.get(
     "WEBUI_AUTH_TRUSTED_EMAIL_HEADER", None
 )
 WEBUI_AUTH_TRUSTED_NAME_HEADER = os.environ.get("WEBUI_AUTH_TRUSTED_NAME_HEADER", None)
+WEBUI_AUTH_TRUSTED_GROUPS_HEADER = os.environ.get(
+    "WEBUI_AUTH_TRUSTED_GROUPS_HEADER", None
+)
+
 
 BYPASS_MODEL_ACCESS_CONTROL = (
     os.environ.get("BYPASS_MODEL_ACCESS_CONTROL", "False").lower() == "true"

backend/open_webui/functions.py

@@ -28,7 +28,10 @@ from open_webui.socket.main import (
 from open_webui.models.functions import Functions
 from open_webui.models.models import Models
 
-from open_webui.utils.plugin import load_function_module_by_id
+from open_webui.utils.plugin import (
+    load_function_module_by_id,
+    get_function_module_from_cache,
+)
 from open_webui.utils.tools import get_tools
 from open_webui.utils.access_control import has_access
 
@@ -53,12 +56,7 @@ log.setLevel(SRC_LOG_LEVELS["MAIN"])
 
 
 def get_function_module_by_id(request: Request, pipe_id: str):
-    # Check if function is already loaded
-    if pipe_id not in request.app.state.FUNCTIONS:
-        function_module, _, _ = load_function_module_by_id(pipe_id)
-        request.app.state.FUNCTIONS[pipe_id] = function_module
-    else:
-        function_module = request.app.state.FUNCTIONS[pipe_id]
+    function_module, _, _ = get_function_module_from_cache(request, pipe_id)
 
     if hasattr(function_module, "valves") and hasattr(function_module, "Valves"):
         valves = Functions.get_function_valves_by_id(pipe_id)

backend/open_webui/internal/wrappers.py

@@ -43,7 +43,7 @@ class ReconnectingPostgresqlDatabase(CustomReconnectMixin, PostgresqlDatabase):
 
 
 def register_connection(db_url):
-    db = connect(db_url, unquote_password=True)
+    db = connect(db_url, unquote_user=True, unquote_password=True)
     if isinstance(db, PostgresqlDatabase):
         # Enable autoconnect for SQLite databases, managed by Peewee
         db.autoconnect = True
@@ -51,7 +51,7 @@ def register_connection(db_url):
         log.info("Connected to PostgreSQL database")
 
         # Get the connection details
-        connection = parse(db_url, unquote_password=True)
+        connection = parse(db_url, unquote_user=True, unquote_password=True)
 
         # Use our custom database class that supports reconnection
         db = ReconnectingPostgresqlDatabase(**connection)

backend/open_webui/main.py

@@ -40,6 +40,8 @@ from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import JSONResponse, RedirectResponse
 from fastapi.staticfiles import StaticFiles
 
+from starlette_compress import CompressMiddleware
+
 from starlette.exceptions import HTTPException as StarletteHTTPException
 from starlette.middleware.base import BaseHTTPMiddleware
 from starlette.middleware.sessions import SessionMiddleware
@@ -196,7 +198,11 @@ from open_webui.config import (
     RAG_RERANKING_MODEL_TRUST_REMOTE_CODE,
     RAG_EMBEDDING_ENGINE,
     RAG_EMBEDDING_BATCH_SIZE,
+    RAG_TOP_K,
+    RAG_TOP_K_RERANKER,
     RAG_RELEVANCE_THRESHOLD,
+    RAG_HYBRID_BM25_WEIGHT,
+    RAG_ALLOWED_FILE_EXTENSIONS,
     RAG_FILE_MAX_COUNT,
     RAG_FILE_MAX_SIZE,
     RAG_OPENAI_API_BASE_URL,
@@ -206,15 +212,25 @@ from open_webui.config import (
     CHUNK_OVERLAP,
     CHUNK_SIZE,
     CONTENT_EXTRACTION_ENGINE,
+    DATALAB_MARKER_API_KEY,
+    DATALAB_MARKER_LANGS,
+    DATALAB_MARKER_SKIP_CACHE,
+    DATALAB_MARKER_FORCE_OCR,
+    DATALAB_MARKER_PAGINATE,
+    DATALAB_MARKER_STRIP_EXISTING_OCR,
+    DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION,
+    DATALAB_MARKER_OUTPUT_FORMAT,
+    DATALAB_MARKER_USE_LLM,
+    EXTERNAL_DOCUMENT_LOADER_URL,
+    EXTERNAL_DOCUMENT_LOADER_API_KEY,
     TIKA_SERVER_URL,
     DOCLING_SERVER_URL,
     DOCLING_OCR_ENGINE,
     DOCLING_OCR_LANG,
+    DOCLING_DO_PICTURE_DESCRIPTION,
     DOCUMENT_INTELLIGENCE_ENDPOINT,
     DOCUMENT_INTELLIGENCE_KEY,
     MISTRAL_OCR_API_KEY,
-    RAG_TOP_K,
-    RAG_TOP_K_RERANKER,
     RAG_TEXT_SPLITTER,
     TIKTOKEN_ENCODING_NAME,
     PDF_EXTRACT_IMAGES,
@@ -224,6 +240,7 @@ from open_webui.config import (
     ENABLE_WEB_SEARCH,
     WEB_SEARCH_ENGINE,
     BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL,
+    BYPASS_WEB_SEARCH_WEB_LOADER,
     WEB_SEARCH_RESULT_COUNT,
     WEB_SEARCH_CONCURRENT_REQUESTS,
     WEB_SEARCH_TRUST_ENV,
@@ -291,6 +308,8 @@ from open_webui.config import (
     ENABLE_EVALUATION_ARENA_MODELS,
     USER_PERMISSIONS,
     DEFAULT_USER_ROLE,
+    PENDING_USER_OVERLAY_CONTENT,
+    PENDING_USER_OVERLAY_TITLE,
     DEFAULT_PROMPT_SUGGESTIONS,
     DEFAULT_MODELS,
     DEFAULT_ARENA_MODEL,
@@ -317,6 +336,7 @@ from open_webui.config import (
     LDAP_APP_PASSWORD,
     LDAP_USE_TLS,
     LDAP_CA_CERT_FILE,
+    LDAP_VALIDATE_CERT,
     LDAP_CIPHERS,
     # Misc
     ENV,
@@ -327,6 +347,7 @@ from open_webui.config import (
     DEFAULT_LOCALE,
     OAUTH_PROVIDERS,
     WEBUI_URL,
+    RESPONSE_WATERMARK,
     # Admin
     ENABLE_ADMIN_CHAT_ACCESS,
     ENABLE_ADMIN_EXPORT,
@@ -373,6 +394,7 @@ from open_webui.env import (
     OFFLINE_MODE,
     ENABLE_OTEL,
     EXTERNAL_PWA_MANIFEST_URL,
+    AIOHTTP_CLIENT_SESSION_SSL,
 )
 
 
@@ -573,6 +595,11 @@ app.state.config.DEFAULT_MODELS = DEFAULT_MODELS
 app.state.config.DEFAULT_PROMPT_SUGGESTIONS = DEFAULT_PROMPT_SUGGESTIONS
 app.state.config.DEFAULT_USER_ROLE = DEFAULT_USER_ROLE
 
+app.state.config.PENDING_USER_OVERLAY_CONTENT = PENDING_USER_OVERLAY_CONTENT
+app.state.config.PENDING_USER_OVERLAY_TITLE = PENDING_USER_OVERLAY_TITLE
+
+app.state.config.RESPONSE_WATERMARK = RESPONSE_WATERMARK
+
 app.state.config.USER_PERMISSIONS = USER_PERMISSIONS
 app.state.config.WEBHOOK_URL = WEBHOOK_URL
 app.state.config.BANNERS = WEBUI_BANNERS
@@ -609,6 +636,7 @@ app.state.config.LDAP_SEARCH_BASE = LDAP_SEARCH_BASE
 app.state.config.LDAP_SEARCH_FILTERS = LDAP_SEARCH_FILTERS
 app.state.config.LDAP_USE_TLS = LDAP_USE_TLS
 app.state.config.LDAP_CA_CERT_FILE = LDAP_CA_CERT_FILE
+app.state.config.LDAP_VALIDATE_CERT = LDAP_VALIDATE_CERT
 app.state.config.LDAP_CIPHERS = LDAP_CIPHERS
 
 
@@ -618,8 +646,12 @@ app.state.WEBUI_AUTH_SIGNOUT_REDIRECT_URL = WEBUI_AUTH_SIGNOUT_REDIRECT_URL
 app.state.EXTERNAL_PWA_MANIFEST_URL = EXTERNAL_PWA_MANIFEST_URL
 
 app.state.USER_COUNT = None
+
 app.state.TOOLS = {}
+app.state.TOOL_CONTENTS = {}
+
 app.state.FUNCTIONS = {}
+app.state.FUNCTION_CONTENTS = {}
 
 ########################################
 #
@@ -631,6 +663,8 @@ app.state.FUNCTIONS = {}
 app.state.config.TOP_K = RAG_TOP_K
 app.state.config.TOP_K_RERANKER = RAG_TOP_K_RERANKER
 app.state.config.RELEVANCE_THRESHOLD = RAG_RELEVANCE_THRESHOLD
+app.state.config.HYBRID_BM25_WEIGHT = RAG_HYBRID_BM25_WEIGHT
+app.state.config.ALLOWED_FILE_EXTENSIONS = RAG_ALLOWED_FILE_EXTENSIONS
 app.state.config.FILE_MAX_SIZE = RAG_FILE_MAX_SIZE
 app.state.config.FILE_MAX_COUNT = RAG_FILE_MAX_COUNT
 
@@ -641,10 +675,24 @@ app.state.config.ENABLE_RAG_HYBRID_SEARCH = ENABLE_RAG_HYBRID_SEARCH
 app.state.config.ENABLE_WEB_LOADER_SSL_VERIFICATION = ENABLE_WEB_LOADER_SSL_VERIFICATION
 
 app.state.config.CONTENT_EXTRACTION_ENGINE = CONTENT_EXTRACTION_ENGINE
+app.state.config.DATALAB_MARKER_API_KEY = DATALAB_MARKER_API_KEY
+app.state.config.DATALAB_MARKER_LANGS = DATALAB_MARKER_LANGS
+app.state.config.DATALAB_MARKER_SKIP_CACHE = DATALAB_MARKER_SKIP_CACHE
+app.state.config.DATALAB_MARKER_FORCE_OCR = DATALAB_MARKER_FORCE_OCR
+app.state.config.DATALAB_MARKER_PAGINATE = DATALAB_MARKER_PAGINATE
+app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR = DATALAB_MARKER_STRIP_EXISTING_OCR
+app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION = (
+    DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION
+)
+app.state.config.DATALAB_MARKER_USE_LLM = DATALAB_MARKER_USE_LLM
+app.state.config.DATALAB_MARKER_OUTPUT_FORMAT = DATALAB_MARKER_OUTPUT_FORMAT
+app.state.config.EXTERNAL_DOCUMENT_LOADER_URL = EXTERNAL_DOCUMENT_LOADER_URL
+app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY = EXTERNAL_DOCUMENT_LOADER_API_KEY
 app.state.config.TIKA_SERVER_URL = TIKA_SERVER_URL
 app.state.config.DOCLING_SERVER_URL = DOCLING_SERVER_URL
 app.state.config.DOCLING_OCR_ENGINE = DOCLING_OCR_ENGINE
 app.state.config.DOCLING_OCR_LANG = DOCLING_OCR_LANG
+app.state.config.DOCLING_DO_PICTURE_DESCRIPTION = DOCLING_DO_PICTURE_DESCRIPTION
 app.state.config.DOCUMENT_INTELLIGENCE_ENDPOINT = DOCUMENT_INTELLIGENCE_ENDPOINT
 app.state.config.DOCUMENT_INTELLIGENCE_KEY = DOCUMENT_INTELLIGENCE_KEY
 app.state.config.MISTRAL_OCR_API_KEY = MISTRAL_OCR_API_KEY
@@ -688,6 +736,7 @@ app.state.config.WEB_SEARCH_TRUST_ENV = WEB_SEARCH_TRUST_ENV
 app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL = (
     BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL
 )
+app.state.config.BYPASS_WEB_SEARCH_WEB_LOADER = BYPASS_WEB_SEARCH_WEB_LOADER
 
 app.state.config.ENABLE_GOOGLE_DRIVE_INTEGRATION = ENABLE_GOOGLE_DRIVE_INTEGRATION
 app.state.config.ENABLE_ONEDRIVE_INTEGRATION = ENABLE_ONEDRIVE_INTEGRATION
@@ -940,6 +989,7 @@ class RedirectMiddleware(BaseHTTPMiddleware):
 
 
 # Add the middleware to the app
+app.add_middleware(CompressMiddleware)
 app.add_middleware(RedirectMiddleware)
 app.add_middleware(SecurityHeadersMiddleware)
 
@@ -1167,11 +1217,12 @@ async def chat_completion(
             "chat_id": form_data.pop("chat_id", None),
             "message_id": form_data.pop("id", None),
             "session_id": form_data.pop("session_id", None),
+            "filter_ids": form_data.pop("filter_ids", []),
             "tool_ids": form_data.get("tool_ids", None),
             "tool_servers": form_data.pop("tool_servers", None),
             "files": form_data.get("files", None),
-            "features": form_data.get("features", None),
-            "variables": form_data.get("variables", None),
+            "features": form_data.get("features", {}),
+            "variables": form_data.get("variables", {}),
             "model": model,
             "direct": model_item.get("direct", False),
             **(
@@ -1395,6 +1446,11 @@ async def get_app_config(request: Request):
                     "sharepoint_url": ONEDRIVE_SHAREPOINT_URL.value,
                     "sharepoint_tenant_id": ONEDRIVE_SHAREPOINT_TENANT_ID.value,
                 },
+                "ui": {
+                    "pending_user_overlay_title": app.state.config.PENDING_USER_OVERLAY_TITLE,
+                    "pending_user_overlay_content": app.state.config.PENDING_USER_OVERLAY_CONTENT,
+                    "response_watermark": app.state.config.RESPONSE_WATERMARK,
+                },
                 "license_metadata": app.state.LICENSE_METADATA,
                 **(
                     {
@@ -1446,7 +1502,8 @@ async def get_app_latest_release_version(user=Depends(get_verified_user)):
         timeout = aiohttp.ClientTimeout(total=1)
         async with aiohttp.ClientSession(timeout=timeout, trust_env=True) as session:
             async with session.get(
-                "https://api.github.com/repos/open-webui/open-webui/releases/latest"
+                "https://api.github.com/repos/open-webui/open-webui/releases/latest",
+                ssl=AIOHTTP_CLIENT_SESSION_SSL,
             ) as response:
                 response.raise_for_status()
                 data = await response.json()

backend/open_webui/models/auths.py

@@ -129,12 +129,16 @@ class AuthsTable:
 
     def authenticate_user(self, email: str, password: str) -> Optional[UserModel]:
         log.info(f"authenticate_user: {email}")
+
+        user = Users.get_user_by_email(email)
+        if not user:
+            return None
+
         try:
             with get_db() as db:
-                auth = db.query(Auth).filter_by(email=email, active=True).first()
+                auth = db.query(Auth).filter_by(id=user.id, active=True).first()
                 if auth:
                     if verify_password(password, auth.password):
-                        user = Users.get_user_by_id(auth.id)
                         return user
                     else:
                         return None
@@ -155,8 +159,8 @@ class AuthsTable:
         except Exception:
             return False
 
-    def authenticate_user_by_trusted_header(self, email: str) -> Optional[UserModel]:
-        log.info(f"authenticate_user_by_trusted_header: {email}")
+    def authenticate_user_by_email(self, email: str) -> Optional[UserModel]:
+        log.info(f"authenticate_user_by_email: {email}")
         try:
             with get_db() as db:
                 auth = db.query(Auth).filter_by(email=email, active=True).first()

backend/open_webui/models/chats.py

@@ -377,22 +377,47 @@ class ChatTable:
             return False
 
     def get_archived_chat_list_by_user_id(
-        self, user_id: str, skip: int = 0, limit: int = 50
+        self,
+        user_id: str,
+        filter: Optional[dict] = None,
+        skip: int = 0,
+        limit: int = 50,
     ) -> list[ChatModel]:
+
         with get_db() as db:
-            all_chats = (
-                db.query(Chat)
-                .filter_by(user_id=user_id, archived=True)
-                .order_by(Chat.updated_at.desc())
-                # .limit(limit).offset(skip)
-                .all()
-            )
+            query = db.query(Chat).filter_by(user_id=user_id, archived=True)
+
+            if filter:
+                query_key = filter.get("query")
+                if query_key:
+                    query = query.filter(Chat.title.ilike(f"%{query_key}%"))
+
+                order_by = filter.get("order_by")
+                direction = filter.get("direction")
+
+                if order_by and direction and getattr(Chat, order_by):
+                    if direction.lower() == "asc":
+                        query = query.order_by(getattr(Chat, order_by).asc())
+                    elif direction.lower() == "desc":
+                        query = query.order_by(getattr(Chat, order_by).desc())
+                    else:
+                        raise ValueError("Invalid direction for ordering")
+            else:
+                query = query.order_by(Chat.updated_at.desc())
+
+            if skip:
+                query = query.offset(skip)
+            if limit:
+                query = query.limit(limit)
+
+            all_chats = query.all()
             return [ChatModel.model_validate(chat) for chat in all_chats]
 
     def get_chat_list_by_user_id(
         self,
         user_id: str,
         include_archived: bool = False,
+        filter: Optional[dict] = None,
         skip: int = 0,
         limit: int = 50,
     ) -> list[ChatModel]:
@@ -401,7 +426,23 @@ class ChatTable:
             if not include_archived:
                 query = query.filter_by(archived=False)
 
-            query = query.order_by(Chat.updated_at.desc())
+            if filter:
+                query_key = filter.get("query")
+                if query_key:
+                    query = query.filter(Chat.title.ilike(f"%{query_key}%"))
+
+                order_by = filter.get("order_by")
+                direction = filter.get("direction")
+
+                if order_by and direction and getattr(Chat, order_by):
+                    if direction.lower() == "asc":
+                        query = query.order_by(getattr(Chat, order_by).asc())
+                    elif direction.lower() == "desc":
+                        query = query.order_by(getattr(Chat, order_by).desc())
+                    else:
+                        raise ValueError("Invalid direction for ordering")
+            else:
+                query = query.order_by(Chat.updated_at.desc())
 
             if skip:
                 query = query.offset(skip)
@@ -542,7 +583,9 @@ class ChatTable:
         search_text = search_text.lower().strip()
 
         if not search_text:
-            return self.get_chat_list_by_user_id(user_id, include_archived, skip, limit)
+            return self.get_chat_list_by_user_id(
+                user_id, include_archived, filter={}, skip=skip, limit=limit
+            )
 
         search_text_words = search_text.split(" ")
 

backend/open_webui/models/functions.py

@@ -108,6 +108,54 @@ class FunctionsTable:
             log.exception(f"Error creating a new function: {e}")
             return None
 
+    def sync_functions(
+        self, user_id: str, functions: list[FunctionModel]
+    ) -> list[FunctionModel]:
+        # Synchronize functions for a user by updating existing ones, inserting new ones, and removing those that are no longer present.
+        try:
+            with get_db() as db:
+                # Get existing functions
+                existing_functions = db.query(Function).all()
+                existing_ids = {func.id for func in existing_functions}
+
+                # Prepare a set of new function IDs
+                new_function_ids = {func.id for func in functions}
+
+                # Update or insert functions
+                for func in functions:
+                    if func.id in existing_ids:
+                        db.query(Function).filter_by(id=func.id).update(
+                            {
+                                **func.model_dump(),
+                                "user_id": user_id,
+                                "updated_at": int(time.time()),
+                            }
+                        )
+                    else:
+                        new_func = Function(
+                            **{
+                                **func.model_dump(),
+                                "user_id": user_id,
+                                "updated_at": int(time.time()),
+                            }
+                        )
+                        db.add(new_func)
+
+                # Remove functions that are no longer present
+                for func in existing_functions:
+                    if func.id not in new_function_ids:
+                        db.delete(func)
+
+                db.commit()
+
+                return [
+                    FunctionModel.model_validate(func)
+                    for func in db.query(Function).all()
+                ]
+        except Exception as e:
+            log.exception(f"Error syncing functions for user {user_id}: {e}")
+            return []
+
     def get_function_by_id(self, id: str) -> Optional[FunctionModel]:
         try:
             with get_db() as db:

backend/open_webui/models/groups.py

@@ -207,5 +207,43 @@ class GroupTable:
             except Exception:
                 return False
 
+    def sync_user_groups_by_group_names(
+        self, user_id: str, group_names: list[str]
+    ) -> bool:
+        with get_db() as db:
+            try:
+                groups = db.query(Group).filter(Group.name.in_(group_names)).all()
+                group_ids = [group.id for group in groups]
+
+                # Remove user from groups not in the new list
+                existing_groups = self.get_groups_by_member_id(user_id)
+
+                for group in existing_groups:
+                    if group.id not in group_ids:
+                        group.user_ids.remove(user_id)
+                        db.query(Group).filter_by(id=group.id).update(
+                            {
+                                "user_ids": group.user_ids,
+                                "updated_at": int(time.time()),
+                            }
+                        )
+
+                # Add user to new groups
+                for group in groups:
+                    if user_id not in group.user_ids:
+                        group.user_ids.append(user_id)
+                        db.query(Group).filter_by(id=group.id).update(
+                            {
+                                "user_ids": group.user_ids,
+                                "updated_at": int(time.time()),
+                            }
+                        )
+
+                db.commit()
+                return True
+            except Exception as e:
+                log.exception(e)
+                return False
+
 
 Groups = GroupTable()

backend/open_webui/retrieval/loaders/datalab_marker.py

@@ -0,0 +1,251 @@
+import os
+import time
+import requests
+import logging
+import json
+from typing import List, Optional
+from langchain_core.documents import Document
+from fastapi import HTTPException, status
+
+log = logging.getLogger(__name__)
+
+
+class DatalabMarkerLoader:
+    def __init__(
+        self,
+        file_path: str,
+        api_key: str,
+        langs: Optional[str] = None,
+        use_llm: bool = False,
+        skip_cache: bool = False,
+        force_ocr: bool = False,
+        paginate: bool = False,
+        strip_existing_ocr: bool = False,
+        disable_image_extraction: bool = False,
+        output_format: str = None,
+    ):
+        self.file_path = file_path
+        self.api_key = api_key
+        self.langs = langs
+        self.use_llm = use_llm
+        self.skip_cache = skip_cache
+        self.force_ocr = force_ocr
+        self.paginate = paginate
+        self.strip_existing_ocr = strip_existing_ocr
+        self.disable_image_extraction = disable_image_extraction
+        self.output_format = output_format
+
+    def _get_mime_type(self, filename: str) -> str:
+        ext = filename.rsplit(".", 1)[-1].lower()
+        mime_map = {
+            "pdf": "application/pdf",
+            "xls": "application/vnd.ms-excel",
+            "xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+            "ods": "application/vnd.oasis.opendocument.spreadsheet",
+            "doc": "application/msword",
+            "docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+            "odt": "application/vnd.oasis.opendocument.text",
+            "ppt": "application/vnd.ms-powerpoint",
+            "pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+            "odp": "application/vnd.oasis.opendocument.presentation",
+            "html": "text/html",
+            "epub": "application/epub+zip",
+            "png": "image/png",
+            "jpeg": "image/jpeg",
+            "jpg": "image/jpeg",
+            "webp": "image/webp",
+            "gif": "image/gif",
+            "tiff": "image/tiff",
+        }
+        return mime_map.get(ext, "application/octet-stream")
+
+    def check_marker_request_status(self, request_id: str) -> dict:
+        url = f"https://www.datalab.to/api/v1/marker/{request_id}"
+        headers = {"X-Api-Key": self.api_key}
+        try:
+            response = requests.get(url, headers=headers)
+            response.raise_for_status()
+            result = response.json()
+            log.info(f"Marker API status check for request {request_id}: {result}")
+            return result
+        except requests.HTTPError as e:
+            log.error(f"Error checking Marker request status: {e}")
+            raise HTTPException(
+                status.HTTP_502_BAD_GATEWAY,
+                detail=f"Failed to check Marker request: {e}",
+            )
+        except ValueError as e:
+            log.error(f"Invalid JSON checking Marker request: {e}")
+            raise HTTPException(
+                status.HTTP_502_BAD_GATEWAY, detail=f"Invalid JSON: {e}"
+            )
+
+    def load(self) -> List[Document]:
+        url = "https://www.datalab.to/api/v1/marker"
+        filename = os.path.basename(self.file_path)
+        mime_type = self._get_mime_type(filename)
+        headers = {"X-Api-Key": self.api_key}
+
+        form_data = {
+            "langs": self.langs,
+            "use_llm": str(self.use_llm).lower(),
+            "skip_cache": str(self.skip_cache).lower(),
+            "force_ocr": str(self.force_ocr).lower(),
+            "paginate": str(self.paginate).lower(),
+            "strip_existing_ocr": str(self.strip_existing_ocr).lower(),
+            "disable_image_extraction": str(self.disable_image_extraction).lower(),
+            "output_format": self.output_format,
+        }
+
+        log.info(
+            f"Datalab Marker POST request parameters: {{'filename': '{filename}', 'mime_type': '{mime_type}', **{form_data}}}"
+        )
+
+        try:
+            with open(self.file_path, "rb") as f:
+                files = {"file": (filename, f, mime_type)}
+                response = requests.post(
+                    url, data=form_data, files=files, headers=headers
+                )
+                response.raise_for_status()
+                result = response.json()
+        except FileNotFoundError:
+            raise HTTPException(
+                status.HTTP_404_NOT_FOUND, detail=f"File not found: {self.file_path}"
+            )
+        except requests.HTTPError as e:
+            raise HTTPException(
+                status.HTTP_400_BAD_REQUEST,
+                detail=f"Datalab Marker request failed: {e}",
+            )
+        except ValueError as e:
+            raise HTTPException(
+                status.HTTP_502_BAD_GATEWAY, detail=f"Invalid JSON response: {e}"
+            )
+        except Exception as e:
+            raise HTTPException(status.HTTP_500_INTERNAL_SERVER_ERROR, detail=str(e))
+
+        if not result.get("success"):
+            raise HTTPException(
+                status.HTTP_400_BAD_REQUEST,
+                detail=f"Datalab Marker request failed: {result.get('error', 'Unknown error')}",
+            )
+
+        check_url = result.get("request_check_url")
+        request_id = result.get("request_id")
+        if not check_url:
+            raise HTTPException(
+                status.HTTP_502_BAD_GATEWAY, detail="No request_check_url returned."
+            )
+
+        for _ in range(300):  # Up to 10 minutes
+            time.sleep(2)
+            try:
+                poll_response = requests.get(check_url, headers=headers)
+                poll_response.raise_for_status()
+                poll_result = poll_response.json()
+            except (requests.HTTPError, ValueError) as e:
+                raw_body = poll_response.text
+                log.error(f"Polling error: {e}, response body: {raw_body}")
+                raise HTTPException(
+                    status.HTTP_502_BAD_GATEWAY, detail=f"Polling failed: {e}"
+                )
+
+            status_val = poll_result.get("status")
+            success_val = poll_result.get("success")
+
+            if status_val == "complete":
+                summary = {
+                    k: poll_result.get(k)
+                    for k in (
+                        "status",
+                        "output_format",
+                        "success",
+                        "error",
+                        "page_count",
+                        "total_cost",
+                    )
+                }
+                log.info(
+                    f"Marker processing completed successfully: {json.dumps(summary, indent=2)}"
+                )
+                break
+
+            if status_val == "failed" or success_val is False:
+                log.error(
+                    f"Marker poll failed full response: {json.dumps(poll_result, indent=2)}"
+                )
+                error_msg = (
+                    poll_result.get("error")
+                    or "Marker returned failure without error message"
+                )
+                raise HTTPException(
+                    status.HTTP_400_BAD_REQUEST,
+                    detail=f"Marker processing failed: {error_msg}",
+                )
+        else:
+            raise HTTPException(
+                status.HTTP_504_GATEWAY_TIMEOUT, detail="Marker processing timed out"
+            )
+
+        if not poll_result.get("success", False):
+            error_msg = poll_result.get("error") or "Unknown processing error"
+            raise HTTPException(
+                status.HTTP_400_BAD_REQUEST,
+                detail=f"Final processing failed: {error_msg}",
+            )
+
+        content_key = self.output_format.lower()
+        raw_content = poll_result.get(content_key)
+
+        if content_key == "json":
+            full_text = json.dumps(raw_content, indent=2)
+        elif content_key in {"markdown", "html"}:
+            full_text = str(raw_content).strip()
+        else:
+            raise HTTPException(
+                status.HTTP_400_BAD_REQUEST,
+                detail=f"Unsupported output format: {self.output_format}",
+            )
+
+        if not full_text:
+            raise HTTPException(
+                status.HTTP_400_BAD_REQUEST,
+                detail="Datalab Marker returned empty content",
+            )
+
+        marker_output_dir = os.path.join("/app/backend/data/uploads", "marker_output")
+        os.makedirs(marker_output_dir, exist_ok=True)
+
+        file_ext_map = {"markdown": "md", "json": "json", "html": "html"}
+        file_ext = file_ext_map.get(content_key, "txt")
+        output_filename = f"{os.path.splitext(filename)[0]}.{file_ext}"
+        output_path = os.path.join(marker_output_dir, output_filename)
+
+        try:
+            with open(output_path, "w", encoding="utf-8") as f:
+                f.write(full_text)
+            log.info(f"Saved Marker output to: {output_path}")
+        except Exception as e:
+            log.warning(f"Failed to write marker output to disk: {e}")
+
+        metadata = {
+            "source": filename,
+            "output_format": poll_result.get("output_format", self.output_format),
+            "page_count": poll_result.get("page_count", 0),
+            "processed_with_llm": self.use_llm,
+            "request_id": request_id or "",
+        }
+
+        images = poll_result.get("images", {})
+        if images:
+            metadata["image_count"] = len(images)
+            metadata["images"] = json.dumps(list(images.keys()))
+
+        for k, v in metadata.items():
+            if isinstance(v, (dict, list)):
+                metadata[k] = json.dumps(v)
+            elif v is None:
+                metadata[k] = ""
+
+        return [Document(page_content=full_text, metadata=metadata)]

backend/open_webui/retrieval/loaders/external_document.py

@@ -0,0 +1,58 @@
+import requests
+import logging
+from typing import Iterator, List, Union
+
+from langchain_core.document_loaders import BaseLoader
+from langchain_core.documents import Document
+from open_webui.env import SRC_LOG_LEVELS
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["RAG"])
+
+
+class ExternalDocumentLoader(BaseLoader):
+    def __init__(
+        self,
+        file_path,
+        url: str,
+        api_key: str,
+        mime_type=None,
+        **kwargs,
+    ) -> None:
+        self.url = url
+        self.api_key = api_key
+
+        self.file_path = file_path
+        self.mime_type = mime_type
+
+    def load(self) -> list[Document]:
+        with open(self.file_path, "rb") as f:
+            data = f.read()
+
+        headers = {}
+        if self.mime_type is not None:
+            headers["Content-Type"] = self.mime_type
+
+        if self.api_key is not None:
+            headers["Authorization"] = f"Bearer {self.api_key}"
+
+        url = self.url
+        if url.endswith("/"):
+            url = url[:-1]
+
+        r = requests.put(f"{url}/process", data=data, headers=headers)
+
+        if r.ok:
+            res = r.json()
+
+            if res:
+                return [
+                    Document(
+                        page_content=res.get("page_content"),
+                        metadata=res.get("metadata"),
+                    )
+                ]
+            else:
+                raise Exception("Error loading document: No content returned")
+        else:
+            raise Exception(f"Error loading document: {r.status_code} {r.text}")

backend/open_webui/retrieval/loaders/external_web.py

@@ -10,7 +10,7 @@ log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
-class ExternalLoader(BaseLoader):
+class ExternalWebLoader(BaseLoader):
     def __init__(
         self,
         web_paths: Union[str, List[str]],
@@ -32,7 +32,7 @@ class ExternalLoader(BaseLoader):
                 response = requests.post(
                     self.external_url,
                     headers={
-                        "User-Agent": "Open WebUI (https://github.com/open-webui/open-webui) RAG Bot",
+                        "User-Agent": "Open WebUI (https://github.com/open-webui/open-webui) External Web Loader",
                         "Authorization": f"Bearer {self.external_api_key}",
                     },
                     json={

backend/open_webui/retrieval/loaders/main.py

@@ -21,7 +21,11 @@ from langchain_community.document_loaders import (
 )
 from langchain_core.documents import Document
 
+from open_webui.retrieval.loaders.external_document import ExternalDocumentLoader
+
 from open_webui.retrieval.loaders.mistral import MistralLoader
+from open_webui.retrieval.loaders.datalab_marker import DatalabMarkerLoader
+
 
 from open_webui.env import SRC_LOG_LEVELS, GLOBAL_LOG_LEVEL
 
@@ -126,14 +130,12 @@ class TikaLoader:
 
 
 class DoclingLoader:
-    def __init__(
-        self, url, file_path=None, mime_type=None, ocr_engine=None, ocr_lang=None
-    ):
+    def __init__(self, url, file_path=None, mime_type=None, params=None):
         self.url = url.rstrip("/")
         self.file_path = file_path
         self.mime_type = mime_type
-        self.ocr_engine = ocr_engine
-        self.ocr_lang = ocr_lang
+
+        self.params = params or {}
 
     def load(self) -> list[Document]:
         with open(self.file_path, "rb") as f:
@@ -150,11 +152,19 @@ class DoclingLoader:
                 "table_mode": "accurate",
             }
 
-            if self.ocr_engine and self.ocr_lang:
-                params["ocr_engine"] = self.ocr_engine
-                params["ocr_lang"] = [
-                    lang.strip() for lang in self.ocr_lang.split(",") if lang.strip()
-                ]
+            if self.params:
+                if self.params.get("do_picture_classification"):
+                    params["do_picture_classification"] = self.params.get(
+                        "do_picture_classification"
+                    )
+
+                if self.params.get("ocr_engine") and self.params.get("ocr_lang"):
+                    params["ocr_engine"] = self.params.get("ocr_engine")
+                    params["ocr_lang"] = [
+                        lang.strip()
+                        for lang in self.params.get("ocr_lang").split(",")
+                        if lang.strip()
+                    ]
 
             endpoint = f"{self.url}/v1alpha/convert/file"
             r = requests.post(endpoint, files=files, data=params)
@@ -207,7 +217,18 @@ class Loader:
     def _get_loader(self, filename: str, file_content_type: str, file_path: str):
         file_ext = filename.split(".")[-1].lower()
 
-        if self.engine == "tika" and self.kwargs.get("TIKA_SERVER_URL"):
+        if (
+            self.engine == "external"
+            and self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_URL")
+            and self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_API_KEY")
+        ):
+            loader = ExternalDocumentLoader(
+                file_path=file_path,
+                url=self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_URL"),
+                api_key=self.kwargs.get("EXTERNAL_DOCUMENT_LOADER_API_KEY"),
+                mime_type=file_content_type,
+            )
+        elif self.engine == "tika" and self.kwargs.get("TIKA_SERVER_URL"):
             if self._is_text_file(file_ext, file_content_type):
                 loader = TextLoader(file_path, autodetect_encoding=True)
             else:
@@ -217,6 +238,49 @@ class Loader:
                     mime_type=file_content_type,
                     extract_images=self.kwargs.get("PDF_EXTRACT_IMAGES"),
                 )
+        elif (
+            self.engine == "datalab_marker"
+            and self.kwargs.get("DATALAB_MARKER_API_KEY")
+            and file_ext
+            in [
+                "pdf",
+                "xls",
+                "xlsx",
+                "ods",
+                "doc",
+                "docx",
+                "odt",
+                "ppt",
+                "pptx",
+                "odp",
+                "html",
+                "epub",
+                "png",
+                "jpeg",
+                "jpg",
+                "webp",
+                "gif",
+                "tiff",
+            ]
+        ):
+            loader = DatalabMarkerLoader(
+                file_path=file_path,
+                api_key=self.kwargs["DATALAB_MARKER_API_KEY"],
+                langs=self.kwargs.get("DATALAB_MARKER_LANGS"),
+                use_llm=self.kwargs.get("DATALAB_MARKER_USE_LLM", False),
+                skip_cache=self.kwargs.get("DATALAB_MARKER_SKIP_CACHE", False),
+                force_ocr=self.kwargs.get("DATALAB_MARKER_FORCE_OCR", False),
+                paginate=self.kwargs.get("DATALAB_MARKER_PAGINATE", False),
+                strip_existing_ocr=self.kwargs.get(
+                    "DATALAB_MARKER_STRIP_EXISTING_OCR", False
+                ),
+                disable_image_extraction=self.kwargs.get(
+                    "DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION", False
+                ),
+                output_format=self.kwargs.get(
+                    "DATALAB_MARKER_OUTPUT_FORMAT", "markdown"
+                ),
+            )
         elif self.engine == "docling" and self.kwargs.get("DOCLING_SERVER_URL"):
             if self._is_text_file(file_ext, file_content_type):
                 loader = TextLoader(file_path, autodetect_encoding=True)
@@ -225,8 +289,13 @@ class Loader:
                     url=self.kwargs.get("DOCLING_SERVER_URL"),
                     file_path=file_path,
                     mime_type=file_content_type,
-                    ocr_engine=self.kwargs.get("DOCLING_OCR_ENGINE"),
-                    ocr_lang=self.kwargs.get("DOCLING_OCR_LANG"),
+                    params={
+                        "ocr_engine": self.kwargs.get("DOCLING_OCR_ENGINE"),
+                        "ocr_lang": self.kwargs.get("DOCLING_OCR_LANG"),
+                        "do_picture_classification": self.kwargs.get(
+                            "DOCLING_DO_PICTURE_DESCRIPTION"
+                        ),
+                    },
                 )
         elif (
             self.engine == "document_intelligence"
@@ -258,6 +327,15 @@ class Loader:
             loader = MistralLoader(
                 api_key=self.kwargs.get("MISTRAL_OCR_API_KEY"), file_path=file_path
             )
+        elif (
+            self.engine == "external"
+            and self.kwargs.get("MISTRAL_OCR_API_KEY") != ""
+            and file_ext
+            in ["pdf"]  # Mistral OCR currently only supports PDF and images
+        ):
+            loader = MistralLoader(
+                api_key=self.kwargs.get("MISTRAL_OCR_API_KEY"), file_path=file_path
+            )
         else:
             if file_ext == "pdf":
                 loader = PyPDFLoader(

backend/open_webui/retrieval/loaders/mistral.py

@@ -1,8 +1,12 @@
 import requests
+import aiohttp
+import asyncio
 import logging
 import os
 import sys
+import time
 from typing import List, Dict, Any
+from contextlib import asynccontextmanager
 
 from langchain_core.documents import Document
 from open_webui.env import SRC_LOG_LEVELS, GLOBAL_LOG_LEVEL
@@ -14,18 +18,29 @@ log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 class MistralLoader:
     """
+    Enhanced Mistral OCR loader with both sync and async support.
     Loads documents by processing them through the Mistral OCR API.
     """
 
     BASE_API_URL = "https://api.mistral.ai/v1"
 
-    def __init__(self, api_key: str, file_path: str):
+    def __init__(
+        self,
+        api_key: str,
+        file_path: str,
+        timeout: int = 300,  # 5 minutes default
+        max_retries: int = 3,
+        enable_debug_logging: bool = False,
+    ):
         """
-        Initializes the loader.
+        Initializes the loader with enhanced features.
 
         Args:
             api_key: Your Mistral API key.
             file_path: The local path to the PDF file to process.
+            timeout: Request timeout in seconds.
+            max_retries: Maximum number of retry attempts.
+            enable_debug_logging: Enable detailed debug logs.
         """
         if not api_key:
             raise ValueError("API key cannot be empty.")
@@ -34,7 +49,23 @@ class MistralLoader:
 
         self.api_key = api_key
         self.file_path = file_path
-        self.headers = {"Authorization": f"Bearer {self.api_key}"}
+        self.timeout = timeout
+        self.max_retries = max_retries
+        self.debug = enable_debug_logging
+
+        # Pre-compute file info for performance
+        self.file_name = os.path.basename(file_path)
+        self.file_size = os.path.getsize(file_path)
+
+        self.headers = {
+            "Authorization": f"Bearer {self.api_key}",
+            "User-Agent": "OpenWebUI-MistralLoader/2.0",
+        }
+
+    def _debug_log(self, message: str, *args) -> None:
+        """Conditional debug logging for performance."""
+        if self.debug:
+            log.debug(message, *args)
 
     def _handle_response(self, response: requests.Response) -> Dict[str, Any]:
         """Checks response status and returns JSON content."""
@@ -54,24 +85,89 @@ class MistralLoader:
             log.error(f"JSON decode error: {json_err} - Response: {response.text}")
             raise  # Re-raise after logging
 
+    async def _handle_response_async(
+        self, response: aiohttp.ClientResponse
+    ) -> Dict[str, Any]:
+        """Async version of response handling with better error info."""
+        try:
+            response.raise_for_status()
+
+            # Check content type
+            content_type = response.headers.get("content-type", "")
+            if "application/json" not in content_type:
+                if response.status == 204:
+                    return {}
+                text = await response.text()
+                raise ValueError(
+                    f"Unexpected content type: {content_type}, body: {text[:200]}..."
+                )
+
+            return await response.json()
+
+        except aiohttp.ClientResponseError as e:
+            error_text = await response.text() if response else "No response"
+            log.error(f"HTTP {e.status}: {e.message} - Response: {error_text[:500]}")
+            raise
+        except aiohttp.ClientError as e:
+            log.error(f"Client error: {e}")
+            raise
+        except Exception as e:
+            log.error(f"Unexpected error processing response: {e}")
+            raise
+
+    def _retry_request_sync(self, request_func, *args, **kwargs):
+        """Synchronous retry logic with exponential backoff."""
+        for attempt in range(self.max_retries):
+            try:
+                return request_func(*args, **kwargs)
+            except (requests.exceptions.RequestException, Exception) as e:
+                if attempt == self.max_retries - 1:
+                    raise
+
+                wait_time = (2**attempt) + 0.5
+                log.warning(
+                    f"Request failed (attempt {attempt + 1}/{self.max_retries}): {e}. Retrying in {wait_time}s..."
+                )
+                time.sleep(wait_time)
+
+    async def _retry_request_async(self, request_func, *args, **kwargs):
+        """Async retry logic with exponential backoff."""
+        for attempt in range(self.max_retries):
+            try:
+                return await request_func(*args, **kwargs)
+            except (aiohttp.ClientError, asyncio.TimeoutError) as e:
+                if attempt == self.max_retries - 1:
+                    raise
+
+                wait_time = (2**attempt) + 0.5
+                log.warning(
+                    f"Request failed (attempt {attempt + 1}/{self.max_retries}): {e}. Retrying in {wait_time}s..."
+                )
+                await asyncio.sleep(wait_time)
+
     def _upload_file(self) -> str:
-        """Uploads the file to Mistral for OCR processing."""
+        """Uploads the file to Mistral for OCR processing (sync version)."""
         log.info("Uploading file to Mistral API")
         url = f"{self.BASE_API_URL}/files"
         file_name = os.path.basename(self.file_path)
 
-        try:
+        def upload_request():
             with open(self.file_path, "rb") as f:
                 files = {"file": (file_name, f, "application/pdf")}
                 data = {"purpose": "ocr"}
 
-                upload_headers = self.headers.copy()  # Avoid modifying self.headers
-
                 response = requests.post(
-                    url, headers=upload_headers, files=files, data=data
+                    url,
+                    headers=self.headers,
+                    files=files,
+                    data=data,
+                    timeout=self.timeout,
                 )
 
-            response_data = self._handle_response(response)
+            return self._handle_response(response)
+
+        try:
+            response_data = self._retry_request_sync(upload_request)
             file_id = response_data.get("id")
             if not file_id:
                 raise ValueError("File ID not found in upload response.")
@@ -81,16 +177,66 @@ class MistralLoader:
             log.error(f"Failed to upload file: {e}")
             raise
 
+    async def _upload_file_async(self, session: aiohttp.ClientSession) -> str:
+        """Async file upload with streaming for better memory efficiency."""
+        url = f"{self.BASE_API_URL}/files"
+
+        async def upload_request():
+            # Create multipart writer for streaming upload
+            writer = aiohttp.MultipartWriter("form-data")
+
+            # Add purpose field
+            purpose_part = writer.append("ocr")
+            purpose_part.set_content_disposition("form-data", name="purpose")
+
+            # Add file part with streaming
+            file_part = writer.append_payload(
+                aiohttp.streams.FilePayload(
+                    self.file_path,
+                    filename=self.file_name,
+                    content_type="application/pdf",
+                )
+            )
+            file_part.set_content_disposition(
+                "form-data", name="file", filename=self.file_name
+            )
+
+            self._debug_log(
+                f"Uploading file: {self.file_name} ({self.file_size:,} bytes)"
+            )
+
+            async with session.post(
+                url,
+                data=writer,
+                headers=self.headers,
+                timeout=aiohttp.ClientTimeout(total=self.timeout),
+            ) as response:
+                return await self._handle_response_async(response)
+
+        response_data = await self._retry_request_async(upload_request)
+
+        file_id = response_data.get("id")
+        if not file_id:
+            raise ValueError("File ID not found in upload response.")
+
+        log.info(f"File uploaded successfully. File ID: {file_id}")
+        return file_id
+
     def _get_signed_url(self, file_id: str) -> str:
-        """Retrieves a temporary signed URL for the uploaded file."""
+        """Retrieves a temporary signed URL for the uploaded file (sync version)."""
         log.info(f"Getting signed URL for file ID: {file_id}")
         url = f"{self.BASE_API_URL}/files/{file_id}/url"
         params = {"expiry": 1}
         signed_url_headers = {**self.headers, "Accept": "application/json"}
 
+        def url_request():
+            response = requests.get(
+                url, headers=signed_url_headers, params=params, timeout=self.timeout
+            )
+            return self._handle_response(response)
+
         try:
-            response = requests.get(url, headers=signed_url_headers, params=params)
-            response_data = self._handle_response(response)
+            response_data = self._retry_request_sync(url_request)
             signed_url = response_data.get("url")
             if not signed_url:
                 raise ValueError("Signed URL not found in response.")
@@ -100,8 +246,36 @@ class MistralLoader:
             log.error(f"Failed to get signed URL: {e}")
             raise
 
+    async def _get_signed_url_async(
+        self, session: aiohttp.ClientSession, file_id: str
+    ) -> str:
+        """Async signed URL retrieval."""
+        url = f"{self.BASE_API_URL}/files/{file_id}/url"
+        params = {"expiry": 1}
+
+        headers = {**self.headers, "Accept": "application/json"}
+
+        async def url_request():
+            self._debug_log(f"Getting signed URL for file ID: {file_id}")
+            async with session.get(
+                url,
+                headers=headers,
+                params=params,
+                timeout=aiohttp.ClientTimeout(total=self.timeout),
+            ) as response:
+                return await self._handle_response_async(response)
+
+        response_data = await self._retry_request_async(url_request)
+
+        signed_url = response_data.get("url")
+        if not signed_url:
+            raise ValueError("Signed URL not found in response.")
+
+        self._debug_log("Signed URL received successfully")
+        return signed_url
+
     def _process_ocr(self, signed_url: str) -> Dict[str, Any]:
-        """Sends the signed URL to the OCR endpoint for processing."""
+        """Sends the signed URL to the OCR endpoint for processing (sync version)."""
         log.info("Processing OCR via Mistral API")
         url = f"{self.BASE_API_URL}/ocr"
         ocr_headers = {
@@ -118,43 +292,198 @@ class MistralLoader:
             "include_image_base64": False,
         }
 
+        def ocr_request():
+            response = requests.post(
+                url, headers=ocr_headers, json=payload, timeout=self.timeout
+            )
+            return self._handle_response(response)
+
         try:
-            response = requests.post(url, headers=ocr_headers, json=payload)
-            ocr_response = self._handle_response(response)
+            ocr_response = self._retry_request_sync(ocr_request)
             log.info("OCR processing done.")
-            log.debug("OCR response: %s", ocr_response)
+            self._debug_log("OCR response: %s", ocr_response)
             return ocr_response
         except Exception as e:
             log.error(f"Failed during OCR processing: {e}")
             raise
 
+    async def _process_ocr_async(
+        self, session: aiohttp.ClientSession, signed_url: str
+    ) -> Dict[str, Any]:
+        """Async OCR processing with timing metrics."""
+        url = f"{self.BASE_API_URL}/ocr"
+
+        headers = {
+            **self.headers,
+            "Content-Type": "application/json",
+            "Accept": "application/json",
+        }
+
+        payload = {
+            "model": "mistral-ocr-latest",
+            "document": {
+                "type": "document_url",
+                "document_url": signed_url,
+            },
+            "include_image_base64": False,
+        }
+
+        async def ocr_request():
+            log.info("Starting OCR processing via Mistral API")
+            start_time = time.time()
+
+            async with session.post(
+                url,
+                json=payload,
+                headers=headers,
+                timeout=aiohttp.ClientTimeout(total=self.timeout),
+            ) as response:
+                ocr_response = await self._handle_response_async(response)
+
+            processing_time = time.time() - start_time
+            log.info(f"OCR processing completed in {processing_time:.2f}s")
+
+            return ocr_response
+
+        return await self._retry_request_async(ocr_request)
+
     def _delete_file(self, file_id: str) -> None:
-        """Deletes the file from Mistral storage."""
+        """Deletes the file from Mistral storage (sync version)."""
         log.info(f"Deleting uploaded file ID: {file_id}")
         url = f"{self.BASE_API_URL}/files/{file_id}"
-        # No specific Accept header needed, default or Authorization is usually sufficient
 
         try:
-            response = requests.delete(url, headers=self.headers)
-            delete_response = self._handle_response(
-                response
-            )  # Check status, ignore response body unless needed
-            log.info(
-                f"File deleted successfully: {delete_response}"
-            )  # Log the response if available
+            response = requests.delete(url, headers=self.headers, timeout=30)
+            delete_response = self._handle_response(response)
+            log.info(f"File deleted successfully: {delete_response}")
         except Exception as e:
             # Log error but don't necessarily halt execution if deletion fails
             log.error(f"Failed to delete file ID {file_id}: {e}")
-            # Depending on requirements, you might choose to raise the error here
+
+    async def _delete_file_async(
+        self, session: aiohttp.ClientSession, file_id: str
+    ) -> None:
+        """Async file deletion with error tolerance."""
+        try:
+
+            async def delete_request():
+                self._debug_log(f"Deleting file ID: {file_id}")
+                async with session.delete(
+                    url=f"{self.BASE_API_URL}/files/{file_id}",
+                    headers=self.headers,
+                    timeout=aiohttp.ClientTimeout(
+                        total=30
+                    ),  # Shorter timeout for cleanup
+                ) as response:
+                    return await self._handle_response_async(response)
+
+            await self._retry_request_async(delete_request)
+            self._debug_log(f"File {file_id} deleted successfully")
+
+        except Exception as e:
+            # Don't fail the entire process if cleanup fails
+            log.warning(f"Failed to delete file ID {file_id}: {e}")
+
+    @asynccontextmanager
+    async def _get_session(self):
+        """Context manager for HTTP session with optimized settings."""
+        connector = aiohttp.TCPConnector(
+            limit=10,  # Total connection limit
+            limit_per_host=5,  # Per-host connection limit
+            ttl_dns_cache=300,  # DNS cache TTL
+            use_dns_cache=True,
+            keepalive_timeout=30,
+            enable_cleanup_closed=True,
+        )
+
+        async with aiohttp.ClientSession(
+            connector=connector,
+            timeout=aiohttp.ClientTimeout(total=self.timeout),
+            headers={"User-Agent": "OpenWebUI-MistralLoader/2.0"},
+        ) as session:
+            yield session
+
+    def _process_results(self, ocr_response: Dict[str, Any]) -> List[Document]:
+        """Process OCR results into Document objects with enhanced metadata."""
+        pages_data = ocr_response.get("pages")
+        if not pages_data:
+            log.warning("No pages found in OCR response.")
+            return [
+                Document(
+                    page_content="No text content found", metadata={"error": "no_pages"}
+                )
+            ]
+
+        documents = []
+        total_pages = len(pages_data)
+        skipped_pages = 0
+
+        for page_data in pages_data:
+            page_content = page_data.get("markdown")
+            page_index = page_data.get("index")  # API uses 0-based index
+
+            if page_content is not None and page_index is not None:
+                # Clean up content efficiently
+                cleaned_content = (
+                    page_content.strip()
+                    if isinstance(page_content, str)
+                    else str(page_content)
+                )
+
+                if cleaned_content:  # Only add non-empty pages
+                    documents.append(
+                        Document(
+                            page_content=cleaned_content,
+                            metadata={
+                                "page": page_index,  # 0-based index from API
+                                "page_label": page_index
+                                + 1,  # 1-based label for convenience
+                                "total_pages": total_pages,
+                                "file_name": self.file_name,
+                                "file_size": self.file_size,
+                                "processing_engine": "mistral-ocr",
+                            },
+                        )
+                    )
+                else:
+                    skipped_pages += 1
+                    self._debug_log(f"Skipping empty page {page_index}")
+            else:
+                skipped_pages += 1
+                self._debug_log(
+                    f"Skipping page due to missing 'markdown' or 'index'. Data: {page_data}"
+                )
+
+        if skipped_pages > 0:
+            log.info(
+                f"Processed {len(documents)} pages, skipped {skipped_pages} empty/invalid pages"
+            )
+
+        if not documents:
+            # Case where pages existed but none had valid markdown/index
+            log.warning(
+                "OCR response contained pages, but none had valid content/index."
+            )
+            return [
+                Document(
+                    page_content="No valid text content found in document",
+                    metadata={"error": "no_valid_pages", "total_pages": total_pages},
+                )
+            ]
+
+        return documents
 
     def load(self) -> List[Document]:
         """
         Executes the full OCR workflow: upload, get URL, process OCR, delete file.
+        Synchronous version for backward compatibility.
 
         Returns:
             A list of Document objects, one for each page processed.
         """
         file_id = None
+        start_time = time.time()
+
         try:
             # 1. Upload file
             file_id = self._upload_file()
@@ -166,53 +495,30 @@ class MistralLoader:
             ocr_response = self._process_ocr(signed_url)
 
             # 4. Process results
-            pages_data = ocr_response.get("pages")
-            if not pages_data:
-                log.warning("No pages found in OCR response.")
-                return [Document(page_content="No text content found", metadata={})]
+            documents = self._process_results(ocr_response)
 
-            documents = []
-            total_pages = len(pages_data)
-            for page_data in pages_data:
-                page_content = page_data.get("markdown")
-                page_index = page_data.get("index")  # API uses 0-based index
-
-                if page_content is not None and page_index is not None:
-                    documents.append(
-                        Document(
-                            page_content=page_content,
-                            metadata={
-                                "page": page_index,  # 0-based index from API
-                                "page_label": page_index
-                                + 1,  # 1-based label for convenience
-                                "total_pages": total_pages,
-                                # Add other relevant metadata from page_data if available/needed
-                                # e.g., page_data.get('width'), page_data.get('height')
-                            },
-                        )
-                    )
-                else:
-                    log.warning(
-                        f"Skipping page due to missing 'markdown' or 'index'. Data: {page_data}"
-                    )
-
-            if not documents:
-                # Case where pages existed but none had valid markdown/index
-                log.warning(
-                    "OCR response contained pages, but none had valid content/index."
-                )
-                return [
-                    Document(
-                        page_content="No text content found in valid pages", metadata={}
-                    )
-                ]
+            total_time = time.time() - start_time
+            log.info(
+                f"Sync OCR workflow completed in {total_time:.2f}s, produced {len(documents)} documents"
+            )
 
             return documents
 
         except Exception as e:
-            log.error(f"An error occurred during the loading process: {e}")
-            # Return an empty list or a specific error document on failure
-            return [Document(page_content=f"Error during processing: {e}", metadata={})]
+            total_time = time.time() - start_time
+            log.error(
+                f"An error occurred during the loading process after {total_time:.2f}s: {e}"
+            )
+            # Return an error document on failure
+            return [
+                Document(
+                    page_content=f"Error during processing: {e}",
+                    metadata={
+                        "error": "processing_failed",
+                        "file_name": self.file_name,
+                    },
+                )
+            ]
         finally:
             # 5. Delete file (attempt even if prior steps failed after upload)
             if file_id:
@@ -223,3 +529,105 @@ class MistralLoader:
                     log.error(
                         f"Cleanup error: Could not delete file ID {file_id}. Reason: {del_e}"
                     )
+
+    async def load_async(self) -> List[Document]:
+        """
+        Asynchronous OCR workflow execution with optimized performance.
+
+        Returns:
+            A list of Document objects, one for each page processed.
+        """
+        file_id = None
+        start_time = time.time()
+
+        try:
+            async with self._get_session() as session:
+                # 1. Upload file with streaming
+                file_id = await self._upload_file_async(session)
+
+                # 2. Get signed URL
+                signed_url = await self._get_signed_url_async(session, file_id)
+
+                # 3. Process OCR
+                ocr_response = await self._process_ocr_async(session, signed_url)
+
+                # 4. Process results
+                documents = self._process_results(ocr_response)
+
+                total_time = time.time() - start_time
+                log.info(
+                    f"Async OCR workflow completed in {total_time:.2f}s, produced {len(documents)} documents"
+                )
+
+                return documents
+
+        except Exception as e:
+            total_time = time.time() - start_time
+            log.error(f"Async OCR workflow failed after {total_time:.2f}s: {e}")
+            return [
+                Document(
+                    page_content=f"Error during OCR processing: {e}",
+                    metadata={
+                        "error": "processing_failed",
+                        "file_name": self.file_name,
+                    },
+                )
+            ]
+        finally:
+            # 5. Cleanup - always attempt file deletion
+            if file_id:
+                try:
+                    async with self._get_session() as session:
+                        await self._delete_file_async(session, file_id)
+                except Exception as cleanup_error:
+                    log.error(f"Cleanup failed for file ID {file_id}: {cleanup_error}")
+
+    @staticmethod
+    async def load_multiple_async(
+        loaders: List["MistralLoader"],
+    ) -> List[List[Document]]:
+        """
+        Process multiple files concurrently for maximum performance.
+
+        Args:
+            loaders: List of MistralLoader instances
+
+        Returns:
+            List of document lists, one for each loader
+        """
+        if not loaders:
+            return []
+
+        log.info(f"Starting concurrent processing of {len(loaders)} files")
+        start_time = time.time()
+
+        # Process all files concurrently
+        tasks = [loader.load_async() for loader in loaders]
+        results = await asyncio.gather(*tasks, return_exceptions=True)
+
+        # Handle any exceptions in results
+        processed_results = []
+        for i, result in enumerate(results):
+            if isinstance(result, Exception):
+                log.error(f"File {i} failed: {result}")
+                processed_results.append(
+                    [
+                        Document(
+                            page_content=f"Error processing file: {result}",
+                            metadata={
+                                "error": "batch_processing_failed",
+                                "file_index": i,
+                            },
+                        )
+                    ]
+                )
+            else:
+                processed_results.append(result)
+
+        total_time = time.time() - start_time
+        total_docs = sum(len(docs) for docs in processed_results)
+        log.info(
+            f"Batch processing completed in {total_time:.2f}s, produced {total_docs} total documents"
+        )
+
+        return processed_results

backend/open_webui/retrieval/models/base_reranker.py

@@ -0,0 +1,8 @@
+from abc import ABC, abstractmethod
+from typing import Optional, List, Tuple
+
+
+class BaseReranker(ABC):
+    @abstractmethod
+    def predict(self, sentences: List[Tuple[str, str]]) -> Optional[List[float]]:
+        pass

backend/open_webui/retrieval/models/colbert.py

@@ -7,11 +7,13 @@ from colbert.modeling.checkpoint import Checkpoint
 
 from open_webui.env import SRC_LOG_LEVELS
 
+from open_webui.retrieval.models.base_reranker import BaseReranker
+
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
-class ColBERT:
+class ColBERT(BaseReranker):
     def __init__(self, name, **kwargs) -> None:
         log.info("ColBERT: Loading model", name)
         self.device = "cuda" if torch.cuda.is_available() else "cpu"

backend/open_webui/retrieval/models/external.py

@@ -3,12 +3,14 @@ import requests
 from typing import Optional, List, Tuple
 
 from open_webui.env import SRC_LOG_LEVELS
+from open_webui.retrieval.models.base_reranker import BaseReranker
+
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
-class ExternalReranker:
+class ExternalReranker(BaseReranker):
     def __init__(
         self,
         api_key: str,

backend/open_webui/retrieval/utils.py

@@ -12,7 +12,7 @@ from langchain_community.retrievers import BM25Retriever
 from langchain_core.documents import Document
 
 from open_webui.config import VECTOR_DB
-from open_webui.retrieval.vector.connector import VECTOR_DB_CLIENT
+from open_webui.retrieval.vector.factory import VECTOR_DB_CLIENT
 
 from open_webui.models.users import UserModel
 from open_webui.models.files import Files
@@ -116,6 +116,7 @@ def query_doc_with_hybrid_search(
     reranking_function,
     k_reranker: int,
     r: float,
+    hybrid_bm25_weight: float,
 ) -> dict:
     try:
         log.debug(f"query_doc_with_hybrid_search:doc {collection_name}")
@@ -131,9 +132,20 @@ def query_doc_with_hybrid_search(
             top_k=k,
         )
 
-        ensemble_retriever = EnsembleRetriever(
-            retrievers=[bm25_retriever, vector_search_retriever], weights=[0.5, 0.5]
-        )
+        if hybrid_bm25_weight <= 0:
+            ensemble_retriever = EnsembleRetriever(
+                retrievers=[vector_search_retriever], weights=[1.0]
+            )
+        elif hybrid_bm25_weight >= 1:
+            ensemble_retriever = EnsembleRetriever(
+                retrievers=[bm25_retriever], weights=[1.0]
+            )
+        else:
+            ensemble_retriever = EnsembleRetriever(
+                retrievers=[bm25_retriever, vector_search_retriever],
+                weights=[hybrid_bm25_weight, 1.0 - hybrid_bm25_weight],
+            )
+
         compressor = RerankCompressor(
             embedding_function=embedding_function,
             top_n=k_reranker,
@@ -313,6 +325,7 @@ def query_collection_with_hybrid_search(
     reranking_function,
     k_reranker: int,
     r: float,
+    hybrid_bm25_weight: float,
 ) -> dict:
     results = []
     error = False
@@ -346,6 +359,7 @@ def query_collection_with_hybrid_search(
                 reranking_function=reranking_function,
                 k_reranker=k_reranker,
                 r=r,
+                hybrid_bm25_weight=hybrid_bm25_weight,
             )
             return result, None
         except Exception as e:
@@ -433,6 +447,7 @@ def get_sources_from_files(
     reranking_function,
     k_reranker,
     r,
+    hybrid_bm25_weight,
     hybrid_search,
     full_context=False,
 ):
@@ -550,6 +565,7 @@ def get_sources_from_files(
                                     reranking_function=reranking_function,
                                     k_reranker=k_reranker,
                                     r=r,
+                                    hybrid_bm25_weight=hybrid_bm25_weight,
                                 )
                             except Exception as e:
                                 log.debug(
@@ -818,7 +834,9 @@ class RerankCompressor(BaseDocumentCompressor):
             )
             scores = util.cos_sim(query_embedding, document_embedding)[0]
 
-        docs_with_scores = list(zip(documents, scores.tolist()))
+        docs_with_scores = list(
+            zip(documents, scores.tolist() if not isinstance(scores, list) else scores)
+        )
         if self.r_score:
             docs_with_scores = [
                 (d, s) for d, s in docs_with_scores if s >= self.r_score

backend/open_webui/retrieval/vector/connector.py

@@ -1,30 +0,0 @@
-from open_webui.config import VECTOR_DB
-
-if VECTOR_DB == "milvus":
-    from open_webui.retrieval.vector.dbs.milvus import MilvusClient
-
-    VECTOR_DB_CLIENT = MilvusClient()
-elif VECTOR_DB == "qdrant":
-    from open_webui.retrieval.vector.dbs.qdrant import QdrantClient
-
-    VECTOR_DB_CLIENT = QdrantClient()
-elif VECTOR_DB == "opensearch":
-    from open_webui.retrieval.vector.dbs.opensearch import OpenSearchClient
-
-    VECTOR_DB_CLIENT = OpenSearchClient()
-elif VECTOR_DB == "pgvector":
-    from open_webui.retrieval.vector.dbs.pgvector import PgvectorClient
-
-    VECTOR_DB_CLIENT = PgvectorClient()
-elif VECTOR_DB == "elasticsearch":
-    from open_webui.retrieval.vector.dbs.elasticsearch import ElasticsearchClient
-
-    VECTOR_DB_CLIENT = ElasticsearchClient()
-elif VECTOR_DB == "pinecone":
-    from open_webui.retrieval.vector.dbs.pinecone import PineconeClient
-
-    VECTOR_DB_CLIENT = PineconeClient()
-else:
-    from open_webui.retrieval.vector.dbs.chroma import ChromaClient
-
-    VECTOR_DB_CLIENT = ChromaClient()

backend/open_webui/retrieval/vector/dbs/pinecone.py

@@ -1,13 +1,12 @@
 from typing import Optional, List, Dict, Any, Union
 import logging
 import time  # for measuring elapsed time
-from pinecone import ServerlessSpec
+from pinecone import Pinecone, ServerlessSpec
 
 import asyncio  # for async upserts
 import functools  # for partial binding in async tasks
 
 import concurrent.futures  # for parallel batch upserts
-from pinecone.grpc import PineconeGRPC  # use gRPC client for faster upserts
 
 from open_webui.retrieval.vector.main import (
     VectorDBBase,
@@ -47,10 +46,8 @@ class PineconeClient(VectorDBBase):
         self.metric = PINECONE_METRIC
         self.cloud = PINECONE_CLOUD
 
-        # Initialize Pinecone gRPC client for improved performance
-        self.client = PineconeGRPC(
-            api_key=self.api_key, environment=self.environment, cloud=self.cloud
-        )
+        # Initialize Pinecone client for improved performance
+        self.client = Pinecone(api_key=self.api_key)
 
         # Persistent executor for batch operations
         self._executor = concurrent.futures.ThreadPoolExecutor(max_workers=5)
@@ -147,8 +144,8 @@ class PineconeClient(VectorDBBase):
         metadatas = []
 
         for match in matches:
-            metadata = match.get("metadata", {})
-            ids.append(match["id"])
+            metadata = getattr(match, "metadata", {}) or {}
+            ids.append(match.id if hasattr(match, "id") else match["id"])
             documents.append(metadata.get("text", ""))
             metadatas.append(metadata)
 
@@ -174,7 +171,8 @@ class PineconeClient(VectorDBBase):
                 filter={"collection_name": collection_name_with_prefix},
                 include_metadata=False,
             )
-            return len(response.matches) > 0
+            matches = getattr(response, "matches", []) or []
+            return len(matches) > 0
         except Exception as e:
             log.exception(
                 f"Error checking collection '{collection_name_with_prefix}': {e}"
@@ -321,32 +319,6 @@ class PineconeClient(VectorDBBase):
             f"Successfully async upserted {len(points)} vectors in batches into '{collection_name_with_prefix}'"
         )
 
-    def streaming_upsert(self, collection_name: str, items: List[VectorItem]) -> None:
-        """Perform a streaming upsert over gRPC for performance testing."""
-        if not items:
-            log.warning("No items to upsert via streaming")
-            return
-
-        collection_name_with_prefix = self._get_collection_name_with_prefix(
-            collection_name
-        )
-        points = self._create_points(items, collection_name_with_prefix)
-
-        # Open a streaming upsert channel
-        stream = self.index.streaming_upsert()
-        try:
-            for point in points:
-                # send each point over the stream
-                stream.send(point)
-            # close the stream to finalize
-            stream.close()
-            log.info(
-                f"Successfully streamed upsert of {len(points)} vectors into '{collection_name_with_prefix}'"
-            )
-        except Exception as e:
-            log.error(f"Error during streaming upsert: {e}")
-            raise
-
     def search(
         self, collection_name: str, vectors: List[List[Union[float, int]]], limit: int
     ) -> Optional[SearchResult]:
@@ -374,7 +346,8 @@ class PineconeClient(VectorDBBase):
                 filter={"collection_name": collection_name_with_prefix},
             )
 
-            if not query_response.matches:
+            matches = getattr(query_response, "matches", []) or []
+            if not matches:
                 # Return empty result if no matches
                 return SearchResult(
                     ids=[[]],
@@ -384,13 +357,13 @@ class PineconeClient(VectorDBBase):
                 )
 
             # Convert to GetResult format
-            get_result = self._result_to_get_result(query_response.matches)
+            get_result = self._result_to_get_result(matches)
 
             # Calculate normalized distances based on metric
             distances = [
                 [
-                    self._normalize_distance(match.score)
-                    for match in query_response.matches
+                    self._normalize_distance(getattr(match, "score", 0.0))
+                    for match in matches
                 ]
             ]
 
@@ -432,7 +405,8 @@ class PineconeClient(VectorDBBase):
                 include_metadata=True,
             )
 
-            return self._result_to_get_result(query_response.matches)
+            matches = getattr(query_response, "matches", []) or []
+            return self._result_to_get_result(matches)
 
         except Exception as e:
             log.error(f"Error querying collection '{collection_name}': {e}")
@@ -456,7 +430,8 @@ class PineconeClient(VectorDBBase):
                 filter={"collection_name": collection_name_with_prefix},
             )
 
-            return self._result_to_get_result(query_response.matches)
+            matches = getattr(query_response, "matches", []) or []
+            return self._result_to_get_result(matches)
 
         except Exception as e:
             log.error(f"Error getting collection '{collection_name}': {e}")
@@ -516,12 +491,12 @@ class PineconeClient(VectorDBBase):
             raise
 
     def close(self):
-        """Shut down the gRPC channel and thread pool."""
+        """Shut down resources."""
         try:
-            self.client.close()
-            log.info("Pinecone gRPC channel closed.")
+            # The new Pinecone client doesn't need explicit closing
+            pass
         except Exception as e:
-            log.warning(f"Failed to close Pinecone gRPC channel: {e}")
+            log.warning(f"Failed to clean up Pinecone resources: {e}")
         self._executor.shutdown(wait=True)
 
     def __enter__(self):

backend/open_webui/retrieval/vector/dbs/qdrant_multitenancy.py

@@ -0,0 +1,712 @@
+import logging
+from typing import Optional, Tuple
+from urllib.parse import urlparse
+
+import grpc
+from open_webui.config import (
+    QDRANT_API_KEY,
+    QDRANT_GRPC_PORT,
+    QDRANT_ON_DISK,
+    QDRANT_PREFER_GRPC,
+    QDRANT_URI,
+)
+from open_webui.env import SRC_LOG_LEVELS
+from open_webui.retrieval.vector.main import (
+    GetResult,
+    SearchResult,
+    VectorDBBase,
+    VectorItem,
+)
+from qdrant_client import QdrantClient as Qclient
+from qdrant_client.http.exceptions import UnexpectedResponse
+from qdrant_client.http.models import PointStruct
+from qdrant_client.models import models
+
+NO_LIMIT = 999999999
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["RAG"])
+
+
+class QdrantClient(VectorDBBase):
+    def __init__(self):
+        self.collection_prefix = "open-webui"
+        self.QDRANT_URI = QDRANT_URI
+        self.QDRANT_API_KEY = QDRANT_API_KEY
+        self.QDRANT_ON_DISK = QDRANT_ON_DISK
+        self.PREFER_GRPC = QDRANT_PREFER_GRPC
+        self.GRPC_PORT = QDRANT_GRPC_PORT
+
+        if not self.QDRANT_URI:
+            self.client = None
+            return
+
+        # Unified handling for either scheme
+        parsed = urlparse(self.QDRANT_URI)
+        host = parsed.hostname or self.QDRANT_URI
+        http_port = parsed.port or 6333  # default REST port
+
+        if self.PREFER_GRPC:
+            self.client = Qclient(
+                host=host,
+                port=http_port,
+                grpc_port=self.GRPC_PORT,
+                prefer_grpc=self.PREFER_GRPC,
+                api_key=self.QDRANT_API_KEY,
+            )
+        else:
+            self.client = Qclient(url=self.QDRANT_URI, api_key=self.QDRANT_API_KEY)
+
+        # Main collection types for multi-tenancy
+        self.MEMORY_COLLECTION = f"{self.collection_prefix}_memories"
+        self.KNOWLEDGE_COLLECTION = f"{self.collection_prefix}_knowledge"
+        self.FILE_COLLECTION = f"{self.collection_prefix}_files"
+        self.WEB_SEARCH_COLLECTION = f"{self.collection_prefix}_web-search"
+        self.HASH_BASED_COLLECTION = f"{self.collection_prefix}_hash-based"
+
+    def _result_to_get_result(self, points) -> GetResult:
+        ids = []
+        documents = []
+        metadatas = []
+
+        for point in points:
+            payload = point.payload
+            ids.append(point.id)
+            documents.append(payload["text"])
+            metadatas.append(payload["metadata"])
+
+        return GetResult(
+            **{
+                "ids": [ids],
+                "documents": [documents],
+                "metadatas": [metadatas],
+            }
+        )
+
+    def _get_collection_and_tenant_id(self, collection_name: str) -> Tuple[str, str]:
+        """
+        Maps the traditional collection name to multi-tenant collection and tenant ID.
+
+        Returns:
+            tuple: (collection_name, tenant_id)
+        """
+        # Check for user memory collections
+        tenant_id = collection_name
+
+        if collection_name.startswith("user-memory-"):
+            return self.MEMORY_COLLECTION, tenant_id
+
+        # Check for file collections
+        elif collection_name.startswith("file-"):
+            return self.FILE_COLLECTION, tenant_id
+
+        # Check for web search collections
+        elif collection_name.startswith("web-search-"):
+            return self.WEB_SEARCH_COLLECTION, tenant_id
+
+        # Handle hash-based collections (YouTube and web URLs)
+        elif len(collection_name) == 63 and all(
+            c in "0123456789abcdef" for c in collection_name
+        ):
+            return self.HASH_BASED_COLLECTION, tenant_id
+
+        else:
+            return self.KNOWLEDGE_COLLECTION, tenant_id
+
+    def _extract_error_message(self, exception):
+        """
+        Extract error message from either HTTP or gRPC exceptions
+
+        Returns:
+            tuple: (status_code, error_message)
+        """
+        # Check if it's an HTTP exception
+        if isinstance(exception, UnexpectedResponse):
+            try:
+                error_data = exception.structured()
+                error_msg = error_data.get("status", {}).get("error", "")
+                return exception.status_code, error_msg
+            except Exception as inner_e:
+                log.error(f"Failed to parse HTTP error: {inner_e}")
+                return exception.status_code, str(exception)
+
+        # Check if it's a gRPC exception
+        elif isinstance(exception, grpc.RpcError):
+            # Extract status code from gRPC error
+            status_code = None
+            if hasattr(exception, "code") and callable(exception.code):
+                status_code = exception.code().value[0]
+
+            # Extract error message
+            error_msg = str(exception)
+            if "details =" in error_msg:
+                # Parse the details line which contains the actual error message
+                try:
+                    details_line = [
+                        line.strip()
+                        for line in error_msg.split("\n")
+                        if "details =" in line
+                    ][0]
+                    error_msg = details_line.split("details =")[1].strip(' "')
+                except (IndexError, AttributeError):
+                    # Fall back to full message if parsing fails
+                    pass
+
+            return status_code, error_msg
+
+        # For any other type of exception
+        return None, str(exception)
+
+    def _is_collection_not_found_error(self, exception):
+        """
+        Check if the exception is due to collection not found, supporting both HTTP and gRPC
+        """
+        status_code, error_msg = self._extract_error_message(exception)
+
+        # HTTP error (404)
+        if (
+            status_code == 404
+            and "Collection" in error_msg
+            and "doesn't exist" in error_msg
+        ):
+            return True
+
+        # gRPC error (NOT_FOUND status)
+        if (
+            isinstance(exception, grpc.RpcError)
+            and exception.code() == grpc.StatusCode.NOT_FOUND
+        ):
+            return True
+
+        return False
+
+    def _is_dimension_mismatch_error(self, exception):
+        """
+        Check if the exception is due to dimension mismatch, supporting both HTTP and gRPC
+        """
+        status_code, error_msg = self._extract_error_message(exception)
+
+        # Common patterns in both HTTP and gRPC
+        return (
+            "Vector dimension error" in error_msg
+            or "dimensions mismatch" in error_msg
+            or "invalid vector size" in error_msg
+        )
+
+    def _create_multi_tenant_collection_if_not_exists(
+        self, mt_collection_name: str, dimension: int = 384
+    ):
+        """
+        Creates a collection with multi-tenancy configuration if it doesn't exist.
+        Default dimension is set to 384 which corresponds to 'sentence-transformers/all-MiniLM-L6-v2'.
+        When creating collections dynamically (insert/upsert), the actual vector dimensions will be used.
+        """
+        try:
+            # Try to create the collection directly - will fail if it already exists
+            self.client.create_collection(
+                collection_name=mt_collection_name,
+                vectors_config=models.VectorParams(
+                    size=dimension,
+                    distance=models.Distance.COSINE,
+                    on_disk=self.QDRANT_ON_DISK,
+                ),
+                hnsw_config=models.HnswConfigDiff(
+                    payload_m=16,  # Enable per-tenant indexing
+                    m=0,
+                    on_disk=self.QDRANT_ON_DISK,
+                ),
+            )
+
+            # Create tenant ID payload index
+            self.client.create_payload_index(
+                collection_name=mt_collection_name,
+                field_name="tenant_id",
+                field_schema=models.KeywordIndexParams(
+                    type=models.KeywordIndexType.KEYWORD,
+                    is_tenant=True,
+                    on_disk=self.QDRANT_ON_DISK,
+                ),
+                wait=True,
+            )
+
+            log.info(
+                f"Multi-tenant collection {mt_collection_name} created with dimension {dimension}!"
+            )
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            # Check for the specific error indicating collection already exists
+            status_code, error_msg = self._extract_error_message(e)
+
+            # HTTP status code 409 or gRPC ALREADY_EXISTS
+            if (isinstance(e, UnexpectedResponse) and status_code == 409) or (
+                isinstance(e, grpc.RpcError)
+                and e.code() == grpc.StatusCode.ALREADY_EXISTS
+            ):
+                if "already exists" in error_msg:
+                    log.debug(f"Collection {mt_collection_name} already exists")
+                    return
+            # If it's not an already exists error, re-raise
+            raise e
+        except Exception as e:
+            raise e
+
+    def _create_points(self, items: list[VectorItem], tenant_id: str):
+        """
+        Create point structs from vector items with tenant ID.
+        """
+        return [
+            PointStruct(
+                id=item["id"],
+                vector=item["vector"],
+                payload={
+                    "text": item["text"],
+                    "metadata": item["metadata"],
+                    "tenant_id": tenant_id,
+                },
+            )
+            for item in items
+        ]
+
+    def has_collection(self, collection_name: str) -> bool:
+        """
+        Check if a logical collection exists by checking for any points with the tenant ID.
+        """
+        if not self.client:
+            return False
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Create tenant filter
+        tenant_filter = models.FieldCondition(
+            key="tenant_id", match=models.MatchValue(value=tenant_id)
+        )
+
+        try:
+            # Try directly querying - most of the time collection should exist
+            response = self.client.query_points(
+                collection_name=mt_collection,
+                query_filter=models.Filter(must=[tenant_filter]),
+                limit=1,
+            )
+
+            # Collection exists with this tenant ID if there are points
+            return len(response.points) > 0
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            if self._is_collection_not_found_error(e):
+                log.debug(f"Collection {mt_collection} doesn't exist")
+                return False
+            else:
+                # For other API errors, log and return False
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unexpected Qdrant error: {error_msg}")
+                return False
+        except Exception as e:
+            # For any other errors, log and return False
+            log.debug(f"Error checking collection {mt_collection}: {e}")
+            return False
+
+    def delete(
+        self,
+        collection_name: str,
+        ids: Optional[list[str]] = None,
+        filter: Optional[dict] = None,
+    ):
+        """
+        Delete vectors by ID or filter from a collection with tenant isolation.
+        """
+        if not self.client:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Create tenant filter
+        tenant_filter = models.FieldCondition(
+            key="tenant_id", match=models.MatchValue(value=tenant_id)
+        )
+
+        must_conditions = [tenant_filter]
+        should_conditions = []
+
+        if ids:
+            for id_value in ids:
+                should_conditions.append(
+                    models.FieldCondition(
+                        key="metadata.id",
+                        match=models.MatchValue(value=id_value),
+                    ),
+                )
+        elif filter:
+            for key, value in filter.items():
+                must_conditions.append(
+                    models.FieldCondition(
+                        key=f"metadata.{key}",
+                        match=models.MatchValue(value=value),
+                    ),
+                )
+
+        try:
+            # Try to delete directly - most of the time collection should exist
+            update_result = self.client.delete(
+                collection_name=mt_collection,
+                points_selector=models.FilterSelector(
+                    filter=models.Filter(must=must_conditions, should=should_conditions)
+                ),
+            )
+
+            return update_result
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            if self._is_collection_not_found_error(e):
+                log.debug(
+                    f"Collection {mt_collection} doesn't exist, nothing to delete"
+                )
+                return None
+            else:
+                # For other API errors, log and re-raise
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unexpected Qdrant error: {error_msg}")
+                raise
+        except Exception as e:
+            # For non-Qdrant exceptions, re-raise
+            raise
+
+    def search(
+        self, collection_name: str, vectors: list[list[float | int]], limit: int
+    ) -> Optional[SearchResult]:
+        """
+        Search for the nearest neighbor items based on the vectors with tenant isolation.
+        """
+        if not self.client:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Get the vector dimension from the query vector
+        dimension = len(vectors[0]) if vectors and len(vectors) > 0 else None
+
+        try:
+            # Try the search operation directly - most of the time collection should exist
+
+            # Create tenant filter
+            tenant_filter = models.FieldCondition(
+                key="tenant_id", match=models.MatchValue(value=tenant_id)
+            )
+
+            # Ensure vector dimensions match the collection
+            collection_dim = self.client.get_collection(
+                mt_collection
+            ).config.params.vectors.size
+
+            if collection_dim != dimension:
+                if collection_dim < dimension:
+                    vectors = [vector[:collection_dim] for vector in vectors]
+                else:
+                    vectors = [
+                        vector + [0] * (collection_dim - dimension)
+                        for vector in vectors
+                    ]
+
+            # Search with tenant filter
+            prefetch_query = models.Prefetch(
+                filter=models.Filter(must=[tenant_filter]),
+                limit=NO_LIMIT,
+            )
+            query_response = self.client.query_points(
+                collection_name=mt_collection,
+                query=vectors[0],
+                prefetch=prefetch_query,
+                limit=limit,
+            )
+
+            get_result = self._result_to_get_result(query_response.points)
+            return SearchResult(
+                ids=get_result.ids,
+                documents=get_result.documents,
+                metadatas=get_result.metadatas,
+                # qdrant distance is [-1, 1], normalize to [0, 1]
+                distances=[
+                    [(point.score + 1.0) / 2.0 for point in query_response.points]
+                ],
+            )
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            if self._is_collection_not_found_error(e):
+                log.debug(
+                    f"Collection {mt_collection} doesn't exist, search returns None"
+                )
+                return None
+            else:
+                # For other API errors, log and re-raise
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unexpected Qdrant error during search: {error_msg}")
+                raise
+        except Exception as e:
+            # For non-Qdrant exceptions, log and return None
+            log.exception(f"Error searching collection '{collection_name}': {e}")
+            return None
+
+    def query(self, collection_name: str, filter: dict, limit: Optional[int] = None):
+        """
+        Query points with filters and tenant isolation.
+        """
+        if not self.client:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Set default limit if not provided
+        if limit is None:
+            limit = NO_LIMIT
+
+        # Create tenant filter
+        tenant_filter = models.FieldCondition(
+            key="tenant_id", match=models.MatchValue(value=tenant_id)
+        )
+
+        # Create metadata filters
+        field_conditions = []
+        for key, value in filter.items():
+            field_conditions.append(
+                models.FieldCondition(
+                    key=f"metadata.{key}", match=models.MatchValue(value=value)
+                )
+            )
+
+        # Combine tenant filter with metadata filters
+        combined_filter = models.Filter(must=[tenant_filter, *field_conditions])
+
+        try:
+            # Try the query directly - most of the time collection should exist
+            points = self.client.query_points(
+                collection_name=mt_collection,
+                query_filter=combined_filter,
+                limit=limit,
+            )
+
+            return self._result_to_get_result(points.points)
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            if self._is_collection_not_found_error(e):
+                log.debug(
+                    f"Collection {mt_collection} doesn't exist, query returns None"
+                )
+                return None
+            else:
+                # For other API errors, log and re-raise
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unexpected Qdrant error during query: {error_msg}")
+                raise
+        except Exception as e:
+            # For non-Qdrant exceptions, log and re-raise
+            log.exception(f"Error querying collection '{collection_name}': {e}")
+            return None
+
+    def get(self, collection_name: str) -> Optional[GetResult]:
+        """
+        Get all items in a collection with tenant isolation.
+        """
+        if not self.client:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Create tenant filter
+        tenant_filter = models.FieldCondition(
+            key="tenant_id", match=models.MatchValue(value=tenant_id)
+        )
+
+        try:
+            # Try to get points directly - most of the time collection should exist
+            points = self.client.query_points(
+                collection_name=mt_collection,
+                query_filter=models.Filter(must=[tenant_filter]),
+                limit=NO_LIMIT,
+            )
+
+            return self._result_to_get_result(points.points)
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            if self._is_collection_not_found_error(e):
+                log.debug(f"Collection {mt_collection} doesn't exist, get returns None")
+                return None
+            else:
+                # For other API errors, log and re-raise
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unexpected Qdrant error during get: {error_msg}")
+                raise
+        except Exception as e:
+            # For non-Qdrant exceptions, log and return None
+            log.exception(f"Error getting collection '{collection_name}': {e}")
+            return None
+
+    def _handle_operation_with_error_retry(
+        self, operation_name, mt_collection, points, dimension
+    ):
+        """
+        Private helper to handle common error cases for insert and upsert operations.
+
+        Args:
+            operation_name: 'insert' or 'upsert'
+            mt_collection: The multi-tenant collection name
+            points: The vector points to insert/upsert
+            dimension: The dimension of the vectors
+
+        Returns:
+            The operation result (for upsert) or None (for insert)
+        """
+        try:
+            if operation_name == "insert":
+                self.client.upload_points(mt_collection, points)
+                return None
+            else:  # upsert
+                return self.client.upsert(mt_collection, points)
+        except (UnexpectedResponse, grpc.RpcError) as e:
+            # Handle collection not found
+            if self._is_collection_not_found_error(e):
+                log.info(
+                    f"Collection {mt_collection} doesn't exist. Creating it with dimension {dimension}."
+                )
+                # Create collection with correct dimensions from our vectors
+                self._create_multi_tenant_collection_if_not_exists(
+                    mt_collection_name=mt_collection, dimension=dimension
+                )
+                # Try operation again - no need for dimension adjustment since we just created with correct dimensions
+                if operation_name == "insert":
+                    self.client.upload_points(mt_collection, points)
+                    return None
+                else:  # upsert
+                    return self.client.upsert(mt_collection, points)
+
+            # Handle dimension mismatch
+            elif self._is_dimension_mismatch_error(e):
+                # For dimension errors, the collection must exist, so get its configuration
+                mt_collection_info = self.client.get_collection(mt_collection)
+                existing_size = mt_collection_info.config.params.vectors.size
+
+                log.info(
+                    f"Dimension mismatch: Collection {mt_collection} expects {existing_size}, got {dimension}"
+                )
+
+                if existing_size < dimension:
+                    # Truncate vectors to fit
+                    log.info(
+                        f"Truncating vectors from {dimension} to {existing_size} dimensions"
+                    )
+                    points = [
+                        PointStruct(
+                            id=point.id,
+                            vector=point.vector[:existing_size],
+                            payload=point.payload,
+                        )
+                        for point in points
+                    ]
+                elif existing_size > dimension:
+                    # Pad vectors with zeros
+                    log.info(
+                        f"Padding vectors from {dimension} to {existing_size} dimensions with zeros"
+                    )
+                    points = [
+                        PointStruct(
+                            id=point.id,
+                            vector=point.vector
+                            + [0] * (existing_size - len(point.vector)),
+                            payload=point.payload,
+                        )
+                        for point in points
+                    ]
+                # Try operation again with adjusted dimensions
+                if operation_name == "insert":
+                    self.client.upload_points(mt_collection, points)
+                    return None
+                else:  # upsert
+                    return self.client.upsert(mt_collection, points)
+            else:
+                # Not a known error we can handle, log and re-raise
+                _, error_msg = self._extract_error_message(e)
+                log.warning(f"Unhandled Qdrant error: {error_msg}")
+                raise
+        except Exception as e:
+            # For non-Qdrant exceptions, re-raise
+            raise
+
+    def insert(self, collection_name: str, items: list[VectorItem]):
+        """
+        Insert items with tenant ID.
+        """
+        if not self.client or not items:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Get dimensions from the actual vectors
+        dimension = len(items[0]["vector"]) if items else None
+
+        # Create points with tenant ID
+        points = self._create_points(items, tenant_id)
+
+        # Handle the operation with error retry
+        return self._handle_operation_with_error_retry(
+            "insert", mt_collection, points, dimension
+        )
+
+    def upsert(self, collection_name: str, items: list[VectorItem]):
+        """
+        Upsert items with tenant ID.
+        """
+        if not self.client or not items:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        # Get dimensions from the actual vectors
+        dimension = len(items[0]["vector"]) if items else None
+
+        # Create points with tenant ID
+        points = self._create_points(items, tenant_id)
+
+        # Handle the operation with error retry
+        return self._handle_operation_with_error_retry(
+            "upsert", mt_collection, points, dimension
+        )
+
+    def reset(self):
+        """
+        Reset the database by deleting all collections.
+        """
+        if not self.client:
+            return None
+
+        collection_names = self.client.get_collections().collections
+        for collection_name in collection_names:
+            if collection_name.name.startswith(self.collection_prefix):
+                self.client.delete_collection(collection_name=collection_name.name)
+
+    def delete_collection(self, collection_name: str):
+        """
+        Delete a collection.
+        """
+        if not self.client:
+            return None
+
+        # Map to multi-tenant collection and tenant ID
+        mt_collection, tenant_id = self._get_collection_and_tenant_id(collection_name)
+
+        tenant_filter = models.FieldCondition(
+            key="tenant_id", match=models.MatchValue(value=tenant_id)
+        )
+
+        field_conditions = [tenant_filter]
+
+        update_result = self.client.delete(
+            collection_name=mt_collection,
+            points_selector=models.FilterSelector(
+                filter=models.Filter(must=field_conditions)
+            ),
+        )
+
+        if self.client.get_collection(mt_collection).points_count == 0:
+            self.client.delete_collection(mt_collection)
+
+        return update_result

backend/open_webui/retrieval/vector/factory.py

@@ -0,0 +1,55 @@
+from open_webui.retrieval.vector.main import VectorDBBase
+from open_webui.retrieval.vector.type import VectorType
+from open_webui.config import VECTOR_DB, ENABLE_QDRANT_MULTITENANCY_MODE
+
+
+class Vector:
+
+    @staticmethod
+    def get_vector(vector_type: str) -> VectorDBBase:
+        """
+        get vector db instance by vector type
+        """
+        match vector_type:
+            case VectorType.MILVUS:
+                from open_webui.retrieval.vector.dbs.milvus import MilvusClient
+
+                return MilvusClient()
+            case VectorType.QDRANT:
+                if ENABLE_QDRANT_MULTITENANCY_MODE:
+                    from open_webui.retrieval.vector.dbs.qdrant_multitenancy import (
+                        QdrantClient,
+                    )
+
+                    return QdrantClient()
+                else:
+                    from open_webui.retrieval.vector.dbs.qdrant import QdrantClient
+
+                    return QdrantClient()
+            case VectorType.PINECONE:
+                from open_webui.retrieval.vector.dbs.pinecone import PineconeClient
+
+                return PineconeClient()
+            case VectorType.OPENSEARCH:
+                from open_webui.retrieval.vector.dbs.opensearch import OpenSearchClient
+
+                return OpenSearchClient()
+            case VectorType.PGVECTOR:
+                from open_webui.retrieval.vector.dbs.pgvector import PgvectorClient
+
+                return PgvectorClient()
+            case VectorType.ELASTICSEARCH:
+                from open_webui.retrieval.vector.dbs.elasticsearch import (
+                    ElasticsearchClient,
+                )
+
+                return ElasticsearchClient()
+            case VectorType.CHROMA:
+                from open_webui.retrieval.vector.dbs.chroma import ChromaClient
+
+                return ChromaClient()
+            case _:
+                raise ValueError(f"Unsupported vector type: {vector_type}")
+
+
+VECTOR_DB_CLIENT = Vector.get_vector(VECTOR_DB)

backend/open_webui/retrieval/vector/type.py

@@ -0,0 +1,11 @@
+from enum import StrEnum
+
+
+class VectorType(StrEnum):
+    MILVUS = "milvus"
+    QDRANT = "qdrant"
+    CHROMA = "chroma"
+    PINECONE = "pinecone"
+    ELASTICSEARCH = "elasticsearch"
+    OPENSEARCH = "opensearch"
+    PGVECTOR = "pgvector"

backend/open_webui/retrieval/web/searchapi.py

@@ -42,7 +42,9 @@ def search_searchapi(
         results = get_filtered_results(results, filter_list)
     return [
         SearchResult(
-            link=result["link"], title=result["title"], snippet=result["snippet"]
+            link=result["link"],
+            title=result.get("title"),
+            snippet=result.get("snippet"),
         )
         for result in results[:count]
     ]

backend/open_webui/retrieval/web/serpapi.py

@@ -42,7 +42,9 @@ def search_serpapi(
         results = get_filtered_results(results, filter_list)
     return [
         SearchResult(
-            link=result["link"], title=result["title"], snippet=result["snippet"]
+            link=result["link"],
+            title=result.get("title"),
+            snippet=result.get("snippet"),
         )
         for result in results[:count]
     ]

backend/open_webui/retrieval/web/utils.py

@@ -25,7 +25,7 @@ from langchain_community.document_loaders.firecrawl import FireCrawlLoader
 from langchain_community.document_loaders.base import BaseLoader
 from langchain_core.documents import Document
 from open_webui.retrieval.loaders.tavily import TavilyLoader
-from open_webui.retrieval.loaders.external import ExternalLoader
+from open_webui.retrieval.loaders.external_web import ExternalWebLoader
 from open_webui.constants import ERROR_MESSAGES
 from open_webui.config import (
     ENABLE_RAG_LOCAL_WEB_FETCH,
@@ -39,7 +39,7 @@ from open_webui.config import (
     EXTERNAL_WEB_LOADER_URL,
     EXTERNAL_WEB_LOADER_API_KEY,
 )
-from open_webui.env import SRC_LOG_LEVELS
+from open_webui.env import SRC_LOG_LEVELS, AIOHTTP_CLIENT_SESSION_SSL
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -515,7 +515,8 @@ class SafeWebBaseLoader(WebBaseLoader):
                         kwargs["ssl"] = False
 
                     async with session.get(
-                        url, **(self.requests_kwargs | kwargs)
+                        url,
+                        **(self.requests_kwargs | kwargs),
                     ) as response:
                         if self.raise_for_status:
                             response.raise_for_status()
@@ -628,7 +629,7 @@ def get_web_loader(
         web_loader_args["extract_depth"] = TAVILY_EXTRACT_DEPTH.value
 
     if WEB_LOADER_ENGINE.value == "external":
-        WebLoaderClass = ExternalLoader
+        WebLoaderClass = ExternalWebLoader
         web_loader_args["external_url"] = EXTERNAL_WEB_LOADER_URL.value
         web_loader_args["external_api_key"] = EXTERNAL_WEB_LOADER_API_KEY.value
 

backend/open_webui/routers/audio.py

@@ -7,6 +7,9 @@ from functools import lru_cache
 from pathlib import Path
 from pydub import AudioSegment
 from pydub.silence import split_on_silence
+from concurrent.futures import ThreadPoolExecutor
+from typing import Optional
+
 
 import aiohttp
 import aiofiles
@@ -17,6 +20,7 @@ from fastapi import (
     Depends,
     FastAPI,
     File,
+    Form,
     HTTPException,
     Request,
     UploadFile,
@@ -38,6 +42,7 @@ from open_webui.config import (
 
 from open_webui.constants import ERROR_MESSAGES
 from open_webui.env import (
+    AIOHTTP_CLIENT_SESSION_SSL,
     AIOHTTP_CLIENT_TIMEOUT,
     ENV,
     SRC_LOG_LEVELS,
@@ -49,7 +54,7 @@ from open_webui.env import (
 router = APIRouter()
 
 # Constants
-MAX_FILE_SIZE_MB = 25
+MAX_FILE_SIZE_MB = 20
 MAX_FILE_SIZE = MAX_FILE_SIZE_MB * 1024 * 1024  # Convert MB to bytes
 AZURE_MAX_FILE_SIZE_MB = 200
 AZURE_MAX_FILE_SIZE = AZURE_MAX_FILE_SIZE_MB * 1024 * 1024  # Convert MB to bytes
@@ -71,35 +76,50 @@ from pydub import AudioSegment
 from pydub.utils import mediainfo
 
 
-def get_audio_convert_format(file_path):
-    """Check if the given file needs to be converted to a different format."""
+def is_audio_conversion_required(file_path):
+    """
+    Check if the given audio file needs conversion to mp3.
+    """
+    SUPPORTED_FORMATS = {"flac", "m4a", "mp3", "mp4", "mpeg", "wav", "webm"}
+
     if not os.path.isfile(file_path):
         log.error(f"File not found: {file_path}")
         return False
 
     try:
         info = mediainfo(file_path)
+        codec_name = info.get("codec_name", "").lower()
+        codec_type = info.get("codec_type", "").lower()
+        codec_tag_string = info.get("codec_tag_string", "").lower()
 
+        if codec_name == "aac" and codec_type == "audio" and codec_tag_string == "mp4a":
+            # File is AAC/mp4a audio, recommend mp3 conversion
+            return True
+
+        # If the codec name or file extension is in the supported formats
         if (
-            info.get("codec_name") == "aac"
-            and info.get("codec_type") == "audio"
-            and info.get("codec_tag_string") == "mp4a"
+            codec_name in SUPPORTED_FORMATS
+            or os.path.splitext(file_path)[1][1:].lower() in SUPPORTED_FORMATS
         ):
-            return "mp4"
-        elif info.get("format_name") == "ogg":
-            return "ogg"
+            return False  # Already supported
+
+        return True
     except Exception as e:
         log.error(f"Error getting audio format: {e}")
         return False
 
-    return None
 
-
-def convert_audio_to_wav(file_path, output_path, conversion_type):
-    """Convert MP4/OGG audio file to WAV format."""
-    audio = AudioSegment.from_file(file_path, format=conversion_type)
-    audio.export(output_path, format="wav")
-    log.info(f"Converted {file_path} to {output_path}")
+def convert_audio_to_mp3(file_path):
+    """Convert audio file to mp3 format."""
+    try:
+        output_path = os.path.splitext(file_path)[0] + ".mp3"
+        audio = AudioSegment.from_file(file_path)
+        audio.export(output_path, format="mp3")
+        log.info(f"Converted {file_path} to {output_path}")
+        return output_path
+    except Exception as e:
+        log.error(f"Error converting audio file: {e}")
+        return None
 
 
 def set_faster_whisper_model(model: str, auto_update: bool = False):
@@ -326,6 +346,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                             else {}
                         ),
                     },
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
                 ) as r:
                     r.raise_for_status()
 
@@ -381,6 +402,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                         "Content-Type": "application/json",
                         "xi-api-key": request.app.state.config.TTS_API_KEY,
                     },
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
                 ) as r:
                     r.raise_for_status()
 
@@ -439,6 +461,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                         "X-Microsoft-OutputFormat": output_format,
                     },
                     data=data,
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
                 ) as r:
                     r.raise_for_status()
 
@@ -507,12 +530,13 @@ async def speech(request: Request, user=Depends(get_verified_user)):
         return FileResponse(file_path)
 
 
-def transcribe(request: Request, file_path):
-    log.info(f"transcribe: {file_path}")
+def transcription_handler(request, file_path, metadata):
     filename = os.path.basename(file_path)
     file_dir = os.path.dirname(file_path)
     id = filename.split(".")[0]
 
+    metadata = metadata or {}
+
     if request.app.state.config.STT_ENGINE == "":
         if request.app.state.faster_whisper_model is None:
             request.app.state.faster_whisper_model = set_faster_whisper_model(
@@ -524,7 +548,7 @@ def transcribe(request: Request, file_path):
             file_path,
             beam_size=5,
             vad_filter=request.app.state.config.WHISPER_VAD_FILTER,
-            language=WHISPER_LANGUAGE,
+            language=metadata.get("language") or WHISPER_LANGUAGE,
         )
         log.info(
             "Detected language '%s' with probability %f"
@@ -542,19 +566,6 @@ def transcribe(request: Request, file_path):
         log.debug(data)
         return data
     elif request.app.state.config.STT_ENGINE == "openai":
-        convert_format = get_audio_convert_format(file_path)
-
-        if convert_format:
-            ext = convert_format.split(".")[-1]
-
-            os.rename(file_path, file_path.replace(".{ext}", f".{convert_format}"))
-            # Convert unsupported audio file to WAV format
-            convert_audio_to_wav(
-                file_path.replace(".{ext}", f".{convert_format}"),
-                file_path,
-                convert_format,
-            )
-
         r = None
         try:
             r = requests.post(
@@ -563,7 +574,14 @@ def transcribe(request: Request, file_path):
                     "Authorization": f"Bearer {request.app.state.config.STT_OPENAI_API_KEY}"
                 },
                 files={"file": (filename, open(file_path, "rb"))},
-                data={"model": request.app.state.config.STT_MODEL},
+                data={
+                    "model": request.app.state.config.STT_MODEL,
+                    **(
+                        {"language": metadata.get("language")}
+                        if metadata.get("language")
+                        else {}
+                    ),
+                },
             )
 
             r.raise_for_status()
@@ -771,41 +789,135 @@ def transcribe(request: Request, file_path):
             )
 
 
+def transcribe(request: Request, file_path: str, metadata: Optional[dict] = None):
+    log.info(f"transcribe: {file_path} {metadata}")
+
+    if is_audio_conversion_required(file_path):
+        file_path = convert_audio_to_mp3(file_path)
+
+    try:
+        file_path = compress_audio(file_path)
+    except Exception as e:
+        log.exception(e)
+
+    # Always produce a list of chunk paths (could be one entry if small)
+    try:
+        chunk_paths = split_audio(file_path, MAX_FILE_SIZE)
+        print(f"Chunk paths: {chunk_paths}")
+    except Exception as e:
+        log.exception(e)
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=ERROR_MESSAGES.DEFAULT(e),
+        )
+
+    results = []
+    try:
+        with ThreadPoolExecutor() as executor:
+            # Submit tasks for each chunk_path
+            futures = [
+                executor.submit(transcription_handler, request, chunk_path, metadata)
+                for chunk_path in chunk_paths
+            ]
+            # Gather results as they complete
+            for future in futures:
+                try:
+                    results.append(future.result())
+                except Exception as transcribe_exc:
+                    raise HTTPException(
+                        status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
+                        detail=f"Error transcribing chunk: {transcribe_exc}",
+                    )
+    finally:
+        # Clean up only the temporary chunks, never the original file
+        for chunk_path in chunk_paths:
+            if chunk_path != file_path and os.path.isfile(chunk_path):
+                try:
+                    os.remove(chunk_path)
+                except Exception:
+                    pass
+
+    return {
+        "text": " ".join([result["text"] for result in results]),
+    }
+
+
 def compress_audio(file_path):
     if os.path.getsize(file_path) > MAX_FILE_SIZE:
+        id = os.path.splitext(os.path.basename(file_path))[
+            0
+        ]  # Handles names with multiple dots
         file_dir = os.path.dirname(file_path)
+
         audio = AudioSegment.from_file(file_path)
         audio = audio.set_frame_rate(16000).set_channels(1)  # Compress audio
-        compressed_path = f"{file_dir}/{id}_compressed.opus"
-        audio.export(compressed_path, format="opus", bitrate="32k")
-        log.debug(f"Compressed audio to {compressed_path}")
 
-        if (
-            os.path.getsize(compressed_path) > MAX_FILE_SIZE
-        ):  # Still larger than MAX_FILE_SIZE after compression
-            raise Exception(ERROR_MESSAGES.FILE_TOO_LARGE(size=f"{MAX_FILE_SIZE_MB}MB"))
+        compressed_path = os.path.join(file_dir, f"{id}_compressed.mp3")
+        audio.export(compressed_path, format="mp3", bitrate="32k")
+        # log.debug(f"Compressed audio to {compressed_path}")  # Uncomment if log is defined
+
         return compressed_path
     else:
         return file_path
 
 
+def split_audio(file_path, max_bytes, format="mp3", bitrate="32k"):
+    """
+    Splits audio into chunks not exceeding max_bytes.
+    Returns a list of chunk file paths. If audio fits, returns list with original path.
+    """
+    file_size = os.path.getsize(file_path)
+    if file_size <= max_bytes:
+        return [file_path]  # Nothing to split
+
+    audio = AudioSegment.from_file(file_path)
+    duration_ms = len(audio)
+    orig_size = file_size
+
+    approx_chunk_ms = max(int(duration_ms * (max_bytes / orig_size)) - 1000, 1000)
+    chunks = []
+    start = 0
+    i = 0
+
+    base, _ = os.path.splitext(file_path)
+
+    while start < duration_ms:
+        end = min(start + approx_chunk_ms, duration_ms)
+        chunk = audio[start:end]
+        chunk_path = f"{base}_chunk_{i}.{format}"
+        chunk.export(chunk_path, format=format, bitrate=bitrate)
+
+        # Reduce chunk duration if still too large
+        while os.path.getsize(chunk_path) > max_bytes and (end - start) > 5000:
+            end = start + ((end - start) // 2)
+            chunk = audio[start:end]
+            chunk.export(chunk_path, format=format, bitrate=bitrate)
+
+        if os.path.getsize(chunk_path) > max_bytes:
+            os.remove(chunk_path)
+            raise Exception("Audio chunk cannot be reduced below max file size.")
+
+        chunks.append(chunk_path)
+        start = end
+        i += 1
+
+    return chunks
+
+
 @router.post("/transcriptions")
 def transcription(
     request: Request,
     file: UploadFile = File(...),
+    language: Optional[str] = Form(None),
     user=Depends(get_verified_user),
 ):
     log.info(f"file.content_type: {file.content_type}")
 
-    supported_filetypes = (
-        "audio/mpeg",
-        "audio/wav",
-        "audio/ogg",
-        "audio/x-m4a",
-        "audio/webm",
-    )
-
-    if not file.content_type.startswith(supported_filetypes):
+    SUPPORTED_CONTENT_TYPES = {"video/webm"}  # Extend if you add more video types!
+    if not (
+        file.content_type.startswith("audio/")
+        or file.content_type in SUPPORTED_CONTENT_TYPES
+    ):
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
             detail=ERROR_MESSAGES.FILE_NOT_SUPPORTED,
@@ -826,19 +938,18 @@ def transcription(
             f.write(contents)
 
         try:
-            try:
-                file_path = compress_audio(file_path)
-            except Exception as e:
-                log.exception(e)
+            metadata = None
 
-                raise HTTPException(
-                    status_code=status.HTTP_400_BAD_REQUEST,
-                    detail=ERROR_MESSAGES.DEFAULT(e),
-                )
+            if language:
+                metadata = {"language": language}
+
+            result = transcribe(request, file_path, metadata)
+
+            return {
+                **result,
+                "filename": os.path.basename(file_path),
+            }
 
-            data = transcribe(request, file_path)
-            file_path = file_path.split("/")[-1]
-            return {**data, "filename": file_path}
         except Exception as e:
             log.exception(e)
 

backend/open_webui/routers/auths.py

@@ -19,19 +19,21 @@ from open_webui.models.auths import (
     UserResponse,
 )
 from open_webui.models.users import Users
+from open_webui.models.groups import Groups
 
 from open_webui.constants import ERROR_MESSAGES, WEBHOOK_MESSAGES
 from open_webui.env import (
     WEBUI_AUTH,
     WEBUI_AUTH_TRUSTED_EMAIL_HEADER,
     WEBUI_AUTH_TRUSTED_NAME_HEADER,
+    WEBUI_AUTH_TRUSTED_GROUPS_HEADER,
     WEBUI_AUTH_COOKIE_SAME_SITE,
     WEBUI_AUTH_COOKIE_SECURE,
     WEBUI_AUTH_SIGNOUT_REDIRECT_URL,
     SRC_LOG_LEVELS,
 )
 from fastapi import APIRouter, Depends, HTTPException, Request, status
-from fastapi.responses import RedirectResponse, Response
+from fastapi.responses import RedirectResponse, Response, JSONResponse
 from open_webui.config import OPENID_PROVIDER_URL, ENABLE_OAUTH_SIGNUP, ENABLE_LDAP
 from pydantic import BaseModel
 
@@ -51,7 +53,7 @@ from open_webui.utils.access_control import get_permissions
 
 from typing import Optional, List
 
-from ssl import CERT_REQUIRED, PROTOCOL_TLS
+from ssl import CERT_NONE, CERT_REQUIRED, PROTOCOL_TLS
 
 if ENABLE_LDAP.value:
     from ldap3 import Server, Connection, NONE, Tls
@@ -186,6 +188,9 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
     LDAP_APP_PASSWORD = request.app.state.config.LDAP_APP_PASSWORD
     LDAP_USE_TLS = request.app.state.config.LDAP_USE_TLS
     LDAP_CA_CERT_FILE = request.app.state.config.LDAP_CA_CERT_FILE
+    LDAP_VALIDATE_CERT = (
+        CERT_REQUIRED if request.app.state.config.LDAP_VALIDATE_CERT else CERT_NONE
+    )
     LDAP_CIPHERS = (
         request.app.state.config.LDAP_CIPHERS
         if request.app.state.config.LDAP_CIPHERS
@@ -197,7 +202,7 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
 
     try:
         tls = Tls(
-            validate=CERT_REQUIRED,
+            validate=LDAP_VALIDATE_CERT,
             version=PROTOCOL_TLS,
             ca_certs_file=LDAP_CA_CERT_FILE,
             ciphers=LDAP_CIPHERS,
@@ -296,7 +301,7 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
                         500, detail="Internal error occurred during LDAP user creation."
                     )
 
-            user = Auths.authenticate_user_by_trusted_header(email)
+            user = Auths.authenticate_user_by_email(email)
 
             if user:
                 expires_delta = parse_duration(request.app.state.config.JWT_EXPIRES_IN)
@@ -360,21 +365,29 @@ async def signin(request: Request, response: Response, form_data: SigninForm):
         if WEBUI_AUTH_TRUSTED_EMAIL_HEADER not in request.headers:
             raise HTTPException(400, detail=ERROR_MESSAGES.INVALID_TRUSTED_HEADER)
 
-        trusted_email = request.headers[WEBUI_AUTH_TRUSTED_EMAIL_HEADER].lower()
-        trusted_name = trusted_email
+        email = request.headers[WEBUI_AUTH_TRUSTED_EMAIL_HEADER].lower()
+        name = email
+
         if WEBUI_AUTH_TRUSTED_NAME_HEADER:
-            trusted_name = request.headers.get(
-                WEBUI_AUTH_TRUSTED_NAME_HEADER, trusted_email
-            )
-        if not Users.get_user_by_email(trusted_email.lower()):
+            name = request.headers.get(WEBUI_AUTH_TRUSTED_NAME_HEADER, email)
+
+        if not Users.get_user_by_email(email.lower()):
             await signup(
                 request,
                 response,
-                SignupForm(
-                    email=trusted_email, password=str(uuid.uuid4()), name=trusted_name
-                ),
+                SignupForm(email=email, password=str(uuid.uuid4()), name=name),
             )
-        user = Auths.authenticate_user_by_trusted_header(trusted_email)
+
+        user = Auths.authenticate_user_by_email(email)
+        if WEBUI_AUTH_TRUSTED_GROUPS_HEADER and user and user.role != "admin":
+            group_names = request.headers.get(
+                WEBUI_AUTH_TRUSTED_GROUPS_HEADER, ""
+            ).split(",")
+            group_names = [name.strip() for name in group_names if name.strip()]
+
+            if group_names:
+                Groups.sync_user_groups_by_group_names(user.id, group_names)
+
     elif WEBUI_AUTH == False:
         admin_email = "admin@localhost"
         admin_password = "admin"
@@ -478,10 +491,6 @@ async def signup(request: Request, response: Response, form_data: SignupForm):
             "admin" if user_count == 0 else request.app.state.config.DEFAULT_USER_ROLE
         )
 
-        if user_count == 0:
-            # Disable signup after the first user is created
-            request.app.state.config.ENABLE_SIGNUP = False
-
         # The password passed to bcrypt must be 72 bytes or fewer. If it is longer, it will be truncated before hashing.
         if len(form_data.password.encode("utf-8")) > 72:
             raise HTTPException(
@@ -541,6 +550,10 @@ async def signup(request: Request, response: Response, form_data: SignupForm):
                 user.id, request.app.state.config.USER_PERMISSIONS
             )
 
+            if user_count == 0:
+                # Disable signup after the first user is created
+                request.app.state.config.ENABLE_SIGNUP = False
+
             return {
                 "token": token,
                 "token_type": "Bearer",
@@ -574,9 +587,14 @@ async def signout(request: Request, response: Response):
                             logout_url = openid_data.get("end_session_endpoint")
                             if logout_url:
                                 response.delete_cookie("oauth_id_token")
-                                return RedirectResponse(
+
+                                return JSONResponse(
+                                    status_code=200,
+                                    content={
+                                        "status": True,
+                                        "redirect_url": f"{logout_url}?id_token_hint={oauth_id_token}",
+                                    },
                                     headers=response.headers,
-                                    url=f"{logout_url}?id_token_hint={oauth_id_token}",
                                 )
                         else:
                             raise HTTPException(
@@ -591,12 +609,18 @@ async def signout(request: Request, response: Response):
                 )
 
     if WEBUI_AUTH_SIGNOUT_REDIRECT_URL:
-        return RedirectResponse(
+        return JSONResponse(
+            status_code=200,
+            content={
+                "status": True,
+                "redirect_url": WEBUI_AUTH_SIGNOUT_REDIRECT_URL,
+            },
             headers=response.headers,
-            url=WEBUI_AUTH_SIGNOUT_REDIRECT_URL,
         )
 
-    return {"status": True}
+    return JSONResponse(
+        status_code=200, content={"status": True}, headers=response.headers
+    )
 
 
 ############################
@@ -696,6 +720,9 @@ async def get_admin_config(request: Request, user=Depends(get_admin_user)):
         "ENABLE_CHANNELS": request.app.state.config.ENABLE_CHANNELS,
         "ENABLE_NOTES": request.app.state.config.ENABLE_NOTES,
         "ENABLE_USER_WEBHOOKS": request.app.state.config.ENABLE_USER_WEBHOOKS,
+        "PENDING_USER_OVERLAY_TITLE": request.app.state.config.PENDING_USER_OVERLAY_TITLE,
+        "PENDING_USER_OVERLAY_CONTENT": request.app.state.config.PENDING_USER_OVERLAY_CONTENT,
+        "RESPONSE_WATERMARK": request.app.state.config.RESPONSE_WATERMARK,
     }
 
 
@@ -713,6 +740,9 @@ class AdminConfig(BaseModel):
     ENABLE_CHANNELS: bool
     ENABLE_NOTES: bool
     ENABLE_USER_WEBHOOKS: bool
+    PENDING_USER_OVERLAY_TITLE: Optional[str] = None
+    PENDING_USER_OVERLAY_CONTENT: Optional[str] = None
+    RESPONSE_WATERMARK: Optional[str] = None
 
 
 @router.post("/admin/config")
@@ -750,6 +780,15 @@ async def update_admin_config(
 
     request.app.state.config.ENABLE_USER_WEBHOOKS = form_data.ENABLE_USER_WEBHOOKS
 
+    request.app.state.config.PENDING_USER_OVERLAY_TITLE = (
+        form_data.PENDING_USER_OVERLAY_TITLE
+    )
+    request.app.state.config.PENDING_USER_OVERLAY_CONTENT = (
+        form_data.PENDING_USER_OVERLAY_CONTENT
+    )
+
+    request.app.state.config.RESPONSE_WATERMARK = form_data.RESPONSE_WATERMARK
+
     return {
         "SHOW_ADMIN_DETAILS": request.app.state.config.SHOW_ADMIN_DETAILS,
         "WEBUI_URL": request.app.state.config.WEBUI_URL,
@@ -764,6 +803,9 @@ async def update_admin_config(
         "ENABLE_CHANNELS": request.app.state.config.ENABLE_CHANNELS,
         "ENABLE_NOTES": request.app.state.config.ENABLE_NOTES,
         "ENABLE_USER_WEBHOOKS": request.app.state.config.ENABLE_USER_WEBHOOKS,
+        "PENDING_USER_OVERLAY_TITLE": request.app.state.config.PENDING_USER_OVERLAY_TITLE,
+        "PENDING_USER_OVERLAY_CONTENT": request.app.state.config.PENDING_USER_OVERLAY_CONTENT,
+        "RESPONSE_WATERMARK": request.app.state.config.RESPONSE_WATERMARK,
     }
 
 
@@ -779,6 +821,7 @@ class LdapServerConfig(BaseModel):
     search_filters: str = ""
     use_tls: bool = True
     certificate_path: Optional[str] = None
+    validate_cert: bool = True
     ciphers: Optional[str] = "ALL"
 
 
@@ -796,6 +839,7 @@ async def get_ldap_server(request: Request, user=Depends(get_admin_user)):
         "search_filters": request.app.state.config.LDAP_SEARCH_FILTERS,
         "use_tls": request.app.state.config.LDAP_USE_TLS,
         "certificate_path": request.app.state.config.LDAP_CA_CERT_FILE,
+        "validate_cert": request.app.state.config.LDAP_VALIDATE_CERT,
         "ciphers": request.app.state.config.LDAP_CIPHERS,
     }
 
@@ -831,6 +875,7 @@ async def update_ldap_server(
     request.app.state.config.LDAP_SEARCH_FILTERS = form_data.search_filters
     request.app.state.config.LDAP_USE_TLS = form_data.use_tls
     request.app.state.config.LDAP_CA_CERT_FILE = form_data.certificate_path
+    request.app.state.config.LDAP_VALIDATE_CERT = form_data.validate_cert
     request.app.state.config.LDAP_CIPHERS = form_data.ciphers
 
     return {
@@ -845,6 +890,7 @@ async def update_ldap_server(
         "search_filters": request.app.state.config.LDAP_SEARCH_FILTERS,
         "use_tls": request.app.state.config.LDAP_USE_TLS,
         "certificate_path": request.app.state.config.LDAP_CA_CERT_FILE,
+        "validate_cert": request.app.state.config.LDAP_VALIDATE_CERT,
         "ciphers": request.app.state.config.LDAP_CIPHERS,
     }
 

backend/open_webui/routers/chats.py

@@ -76,17 +76,34 @@ async def delete_all_user_chats(request: Request, user=Depends(get_verified_user
 @router.get("/list/user/{user_id}", response_model=list[ChatTitleIdResponse])
 async def get_user_chat_list_by_user_id(
     user_id: str,
+    page: Optional[int] = None,
+    query: Optional[str] = None,
+    order_by: Optional[str] = None,
+    direction: Optional[str] = None,
     user=Depends(get_admin_user),
-    skip: int = 0,
-    limit: int = 50,
 ):
     if not ENABLE_ADMIN_CHAT_ACCESS:
         raise HTTPException(
             status_code=status.HTTP_401_UNAUTHORIZED,
             detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
         )
+
+    if page is None:
+        page = 1
+
+    limit = 60
+    skip = (page - 1) * limit
+
+    filter = {}
+    if query:
+        filter["query"] = query
+    if order_by:
+        filter["order_by"] = order_by
+    if direction:
+        filter["direction"] = direction
+
     return Chats.get_chat_list_by_user_id(
-        user_id, include_archived=True, skip=skip, limit=limit
+        user_id, include_archived=True, filter=filter, skip=skip, limit=limit
     )
 
 
@@ -194,10 +211,10 @@ async def get_chats_by_folder_id(folder_id: str, user=Depends(get_verified_user)
 ############################
 
 
-@router.get("/pinned", response_model=list[ChatResponse])
+@router.get("/pinned", response_model=list[ChatTitleIdResponse])
 async def get_user_pinned_chats(user=Depends(get_verified_user)):
     return [
-        ChatResponse(**chat.model_dump())
+        ChatTitleIdResponse(**chat.model_dump())
         for chat in Chats.get_pinned_chats_by_user_id(user.id)
     ]
 
@@ -267,9 +284,37 @@ async def get_all_user_chats_in_db(user=Depends(get_admin_user)):
 
 @router.get("/archived", response_model=list[ChatTitleIdResponse])
 async def get_archived_session_user_chat_list(
-    user=Depends(get_verified_user), skip: int = 0, limit: int = 50
+    page: Optional[int] = None,
+    query: Optional[str] = None,
+    order_by: Optional[str] = None,
+    direction: Optional[str] = None,
+    user=Depends(get_verified_user),
 ):
-    return Chats.get_archived_chat_list_by_user_id(user.id, skip, limit)
+    if page is None:
+        page = 1
+
+    limit = 60
+    skip = (page - 1) * limit
+
+    filter = {}
+    if query:
+        filter["query"] = query
+    if order_by:
+        filter["order_by"] = order_by
+    if direction:
+        filter["direction"] = direction
+
+    chat_list = [
+        ChatTitleIdResponse(**chat.model_dump())
+        for chat in Chats.get_archived_chat_list_by_user_id(
+            user.id,
+            filter=filter,
+            skip=skip,
+            limit=limit,
+        )
+    ]
+
+    return chat_list
 
 
 ############################

backend/open_webui/routers/evaluations.py

@@ -74,13 +74,17 @@ class FeedbackUserResponse(FeedbackResponse):
 @router.get("/feedbacks/all", response_model=list[FeedbackUserResponse])
 async def get_all_feedbacks(user=Depends(get_admin_user)):
     feedbacks = Feedbacks.get_all_feedbacks()
-    return [
-        FeedbackUserResponse(
-            **feedback.model_dump(),
-            user=UserResponse(**Users.get_user_by_id(feedback.user_id).model_dump()),
+
+    feedback_list = []
+    for feedback in feedbacks:
+        user = Users.get_user_by_id(feedback.user_id)
+        feedback_list.append(
+            FeedbackUserResponse(
+                **feedback.model_dump(),
+                user=UserResponse(**user.model_dump()) if user else None,
+            )
         )
-        for feedback in feedbacks
-    ]
+    return feedback_list
 
 
 @router.delete("/feedbacks/all")
@@ -92,12 +96,7 @@ async def delete_all_feedbacks(user=Depends(get_admin_user)):
 @router.get("/feedbacks/all/export", response_model=list[FeedbackModel])
 async def get_all_feedbacks(user=Depends(get_admin_user)):
     feedbacks = Feedbacks.get_all_feedbacks()
-    return [
-        FeedbackModel(
-            **feedback.model_dump(), user=Users.get_user_by_id(feedback.user_id)
-        )
-        for feedback in feedbacks
-    ]
+    return feedbacks
 
 
 @router.get("/feedbacks/user", response_model=list[FeedbackUserResponse])

backend/open_webui/routers/files.py

@@ -1,6 +1,7 @@
 import logging
 import os
 import uuid
+import json
 from fnmatch import fnmatch
 from pathlib import Path
 from typing import Optional
@@ -10,6 +11,7 @@ from fastapi import (
     APIRouter,
     Depends,
     File,
+    Form,
     HTTPException,
     Request,
     UploadFile,
@@ -84,17 +86,44 @@ def has_access_to_file(
 def upload_file(
     request: Request,
     file: UploadFile = File(...),
-    user=Depends(get_verified_user),
-    file_metadata: dict = None,
+    metadata: Optional[dict | str] = Form(None),
     process: bool = Query(True),
+    internal: bool = False,
+    user=Depends(get_verified_user),
 ):
     log.info(f"file.content_type: {file.content_type}")
 
-    file_metadata = file_metadata if file_metadata else {}
+    if isinstance(metadata, str):
+        try:
+            metadata = json.loads(metadata)
+        except json.JSONDecodeError:
+            raise HTTPException(
+                status_code=status.HTTP_400_BAD_REQUEST,
+                detail=ERROR_MESSAGES.DEFAULT("Invalid metadata format"),
+            )
+    file_metadata = metadata if metadata else {}
+
     try:
         unsanitized_filename = file.filename
         filename = os.path.basename(unsanitized_filename)
 
+        file_extension = os.path.splitext(filename)[1]
+        # Remove the leading dot from the file extension
+        file_extension = file_extension[1:] if file_extension else ""
+
+        if (not internal) and request.app.state.config.ALLOWED_FILE_EXTENSIONS:
+            request.app.state.config.ALLOWED_FILE_EXTENSIONS = [
+                ext for ext in request.app.state.config.ALLOWED_FILE_EXTENSIONS if ext
+            ]
+
+            if file_extension not in request.app.state.config.ALLOWED_FILE_EXTENSIONS:
+                raise HTTPException(
+                    status_code=status.HTTP_400_BAD_REQUEST,
+                    detail=ERROR_MESSAGES.DEFAULT(
+                        f"File type {file_extension} is not allowed"
+                    ),
+                )
+
         # replace filename with uuid
         id = str(uuid.uuid4())
         name = filename
@@ -125,33 +154,26 @@ def upload_file(
         )
         if process:
             try:
+                if file.content_type:
+                    if file.content_type.startswith("audio/") or file.content_type in {
+                        "video/webm"
+                    }:
+                        file_path = Storage.get_file(file_path)
+                        result = transcribe(request, file_path, file_metadata)
 
-                if file.content_type.startswith(
-                    (
-                        "audio/mpeg",
-                        "audio/wav",
-                        "audio/ogg",
-                        "audio/x-m4a",
-                        "audio/webm",
-                        "video/webm",
+                        process_file(
+                            request,
+                            ProcessFileForm(file_id=id, content=result.get("text", "")),
+                            user=user,
+                        )
+                    elif (not file.content_type.startswith(("image/", "video/"))) or (
+                        request.app.state.config.CONTENT_EXTRACTION_ENGINE == "external"
+                    ):
+                        process_file(request, ProcessFileForm(file_id=id), user=user)
+                else:
+                    log.info(
+                        f"File type {file.content_type} is not provided, but trying to process anyway"
                     )
-                ):
-                    file_path = Storage.get_file(file_path)
-                    result = transcribe(request, file_path)
-
-                    process_file(
-                        request,
-                        ProcessFileForm(file_id=id, content=result.get("text", "")),
-                        user=user,
-                    )
-                elif file.content_type not in [
-                    "image/png",
-                    "image/jpeg",
-                    "image/gif",
-                    "video/mp4",
-                    "video/ogg",
-                    "video/quicktime",
-                ]:
                     process_file(request, ProcessFileForm(file_id=id), user=user)
 
                 file_item = Files.get_file_by_id(id=id)
@@ -177,7 +199,7 @@ def upload_file(
         log.exception(e)
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
-            detail=ERROR_MESSAGES.DEFAULT(e),
+            detail=ERROR_MESSAGES.DEFAULT("Error uploading file"),
         )
 
 

backend/open_webui/routers/functions.py

@@ -1,5 +1,8 @@
 import os
+import re
+
 import logging
+import aiohttp
 from pathlib import Path
 from typing import Optional
 
@@ -9,12 +12,18 @@ from open_webui.models.functions import (
     FunctionResponse,
     Functions,
 )
-from open_webui.utils.plugin import load_function_module_by_id, replace_imports
+from open_webui.utils.plugin import (
+    load_function_module_by_id,
+    replace_imports,
+    get_function_module_from_cache,
+)
 from open_webui.config import CACHE_DIR
 from open_webui.constants import ERROR_MESSAGES
 from fastapi import APIRouter, Depends, HTTPException, Request, status
 from open_webui.utils.auth import get_admin_user, get_verified_user
 from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, HttpUrl
+
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MAIN"])
@@ -42,6 +51,97 @@ async def get_functions(user=Depends(get_admin_user)):
     return Functions.get_functions()
 
 
+############################
+# LoadFunctionFromLink
+############################
+
+
+class LoadUrlForm(BaseModel):
+    url: HttpUrl
+
+
+def github_url_to_raw_url(url: str) -> str:
+    # Handle 'tree' (folder) URLs (add main.py at the end)
+    m1 = re.match(r"https://github\.com/([^/]+)/([^/]+)/tree/([^/]+)/(.*)", url)
+    if m1:
+        org, repo, branch, path = m1.groups()
+        return f"https://raw.githubusercontent.com/{org}/{repo}/refs/heads/{branch}/{path.rstrip('/')}/main.py"
+
+    # Handle 'blob' (file) URLs
+    m2 = re.match(r"https://github\.com/([^/]+)/([^/]+)/blob/([^/]+)/(.*)", url)
+    if m2:
+        org, repo, branch, path = m2.groups()
+        return (
+            f"https://raw.githubusercontent.com/{org}/{repo}/refs/heads/{branch}/{path}"
+        )
+
+    # No match; return as-is
+    return url
+
+
+@router.post("/load/url", response_model=Optional[dict])
+async def load_function_from_url(
+    request: Request, form_data: LoadUrlForm, user=Depends(get_admin_user)
+):
+    # NOTE: This is NOT a SSRF vulnerability:
+    # This endpoint is admin-only (see get_admin_user), meant for *trusted* internal use,
+    # and does NOT accept untrusted user input. Access is enforced by authentication.
+
+    url = str(form_data.url)
+    if not url:
+        raise HTTPException(status_code=400, detail="Please enter a valid URL")
+
+    url = github_url_to_raw_url(url)
+    url_parts = url.rstrip("/").split("/")
+
+    file_name = url_parts[-1]
+    function_name = (
+        file_name[:-3]
+        if (
+            file_name.endswith(".py")
+            and (not file_name.startswith(("main.py", "index.py", "__init__.py")))
+        )
+        else url_parts[-2] if len(url_parts) > 1 else "function"
+    )
+
+    try:
+        async with aiohttp.ClientSession() as session:
+            async with session.get(
+                url, headers={"Content-Type": "application/json"}
+            ) as resp:
+                if resp.status != 200:
+                    raise HTTPException(
+                        status_code=resp.status, detail="Failed to fetch the function"
+                    )
+                data = await resp.text()
+                if not data:
+                    raise HTTPException(
+                        status_code=400, detail="No data received from the URL"
+                    )
+        return {
+            "name": function_name,
+            "content": data,
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error importing function: {e}")
+
+
+############################
+# SyncFunctions
+############################
+
+
+class SyncFunctionsForm(FunctionForm):
+    functions: list[FunctionModel] = []
+
+
+@router.post("/sync", response_model=Optional[FunctionModel])
+async def sync_functions(
+    request: Request, form_data: SyncFunctionsForm, user=Depends(get_admin_user)
+):
+    return Functions.sync_functions(user.id, form_data.functions)
+
+
 ############################
 # CreateNewFunction
 ############################
@@ -262,11 +362,9 @@ async def get_function_valves_spec_by_id(
 ):
     function = Functions.get_function_by_id(id)
     if function:
-        if id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[id]
-        else:
-            function_module, function_type, frontmatter = load_function_module_by_id(id)
-            request.app.state.FUNCTIONS[id] = function_module
+        function_module, function_type, frontmatter = get_function_module_from_cache(
+            request, id
+        )
 
         if hasattr(function_module, "Valves"):
             Valves = function_module.Valves
@@ -290,11 +388,9 @@ async def update_function_valves_by_id(
 ):
     function = Functions.get_function_by_id(id)
     if function:
-        if id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[id]
-        else:
-            function_module, function_type, frontmatter = load_function_module_by_id(id)
-            request.app.state.FUNCTIONS[id] = function_module
+        function_module, function_type, frontmatter = get_function_module_from_cache(
+            request, id
+        )
 
         if hasattr(function_module, "Valves"):
             Valves = function_module.Valves
@@ -353,11 +449,9 @@ async def get_function_user_valves_spec_by_id(
 ):
     function = Functions.get_function_by_id(id)
     if function:
-        if id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[id]
-        else:
-            function_module, function_type, frontmatter = load_function_module_by_id(id)
-            request.app.state.FUNCTIONS[id] = function_module
+        function_module, function_type, frontmatter = get_function_module_from_cache(
+            request, id
+        )
 
         if hasattr(function_module, "UserValves"):
             UserValves = function_module.UserValves
@@ -377,11 +471,9 @@ async def update_function_user_valves_by_id(
     function = Functions.get_function_by_id(id)
 
     if function:
-        if id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[id]
-        else:
-            function_module, function_type, frontmatter = load_function_module_by_id(id)
-            request.app.state.FUNCTIONS[id] = function_module
+        function_module, function_type, frontmatter = get_function_module_from_cache(
+            request, id
+        )
 
         if hasattr(function_module, "UserValves"):
             UserValves = function_module.UserValves

backend/open_webui/routers/images.py

@@ -333,10 +333,11 @@ def get_models(request: Request, user=Depends(get_verified_user)):
             return [
                 {"id": "dall-e-2", "name": "DALL·E 2"},
                 {"id": "dall-e-3", "name": "DALL·E 3"},
+                {"id": "gpt-image-1", "name": "GPT-IMAGE 1"},
             ]
         elif request.app.state.config.IMAGE_GENERATION_ENGINE == "gemini":
             return [
-                {"id": "imagen-3-0-generate-002", "name": "imagen-3.0 generate-002"},
+                {"id": "imagen-3.0-generate-002", "name": "imagen-3.0 generate-002"},
             ]
         elif request.app.state.config.IMAGE_GENERATION_ENGINE == "comfyui":
             # TODO - get models from comfyui
@@ -450,7 +451,7 @@ def load_url_image_data(url, headers=None):
         return None
 
 
-def upload_image(request, image_metadata, image_data, content_type, user):
+def upload_image(request, image_data, content_type, metadata, user):
     image_format = mimetypes.guess_extension(content_type)
     file = UploadFile(
         file=io.BytesIO(image_data),
@@ -459,7 +460,7 @@ def upload_image(request, image_metadata, image_data, content_type, user):
             "content-type": content_type,
         },
     )
-    file_item = upload_file(request, file, user, file_metadata=image_metadata)
+    file_item = upload_file(request, file, metadata=metadata, internal=True, user=user)
     url = request.app.url_path_for("get_file_content_by_id", id=file_item.id)
     return url
 
@@ -526,7 +527,7 @@ async def image_generations(
                 else:
                     image_data, content_type = load_b64_image_data(image["b64_json"])
 
-                url = upload_image(request, data, image_data, content_type, user)
+                url = upload_image(request, image_data, content_type, data, user)
                 images.append({"url": url})
             return images
 
@@ -560,7 +561,7 @@ async def image_generations(
                 image_data, content_type = load_b64_image_data(
                     image["bytesBase64Encoded"]
                 )
-                url = upload_image(request, data, image_data, content_type, user)
+                url = upload_image(request, image_data, content_type, data, user)
                 images.append({"url": url})
 
             return images
@@ -611,9 +612,9 @@ async def image_generations(
                 image_data, content_type = load_url_image_data(image["url"], headers)
                 url = upload_image(
                     request,
-                    form_data.model_dump(exclude_none=True),
                     image_data,
                     content_type,
+                    form_data.model_dump(exclude_none=True),
                     user,
                 )
                 images.append({"url": url})
@@ -664,9 +665,9 @@ async def image_generations(
                 image_data, content_type = load_b64_image_data(image)
                 url = upload_image(
                     request,
-                    {**data, "info": res["info"]},
                     image_data,
                     content_type,
+                    {**data, "info": res["info"]},
                     user,
                 )
                 images.append({"url": url})

backend/open_webui/routers/knowledge.py

@@ -10,7 +10,7 @@ from open_webui.models.knowledge import (
     KnowledgeUserResponse,
 )
 from open_webui.models.files import Files, FileModel, FileMetadataResponse
-from open_webui.retrieval.vector.connector import VECTOR_DB_CLIENT
+from open_webui.retrieval.vector.factory import VECTOR_DB_CLIENT
 from open_webui.routers.retrieval import (
     process_file,
     ProcessFileForm,

backend/open_webui/routers/memories.py

@@ -4,7 +4,7 @@ import logging
 from typing import Optional
 
 from open_webui.models.memories import Memories, MemoryModel
-from open_webui.retrieval.vector.connector import VECTOR_DB_CLIENT
+from open_webui.retrieval.vector.factory import VECTOR_DB_CLIENT
 from open_webui.utils.auth import get_verified_user
 from open_webui.env import SRC_LOG_LEVELS
 

backend/open_webui/routers/ollama.py

@@ -9,6 +9,8 @@ import os
 import random
 import re
 import time
+from datetime import datetime
+
 from typing import Optional, Union
 from urllib.parse import urlparse
 import aiohttp
@@ -300,6 +302,22 @@ async def update_config(
     }
 
 
+def merge_ollama_models_lists(model_lists):
+    merged_models = {}
+
+    for idx, model_list in enumerate(model_lists):
+        if model_list is not None:
+            for model in model_list:
+                id = model["model"]
+                if id not in merged_models:
+                    model["urls"] = [idx]
+                    merged_models[id] = model
+                else:
+                    merged_models[id]["urls"].append(idx)
+
+    return list(merged_models.values())
+
+
 @cached(ttl=1)
 async def get_all_models(request: Request, user: UserModel = None):
     log.info("get_all_models()")
@@ -340,6 +358,8 @@ async def get_all_models(request: Request, user: UserModel = None):
                     ),  # Legacy support
                 )
 
+                connection_type = api_config.get("connection_type", "local")
+
                 prefix_id = api_config.get("prefix_id", None)
                 tags = api_config.get("tags", [])
                 model_ids = api_config.get("model_ids", [])
@@ -352,31 +372,18 @@ async def get_all_models(request: Request, user: UserModel = None):
                         )
                     )
 
-                if prefix_id:
-                    for model in response.get("models", []):
+                for model in response.get("models", []):
+                    if prefix_id:
                         model["model"] = f"{prefix_id}.{model['model']}"
 
-                if tags:
-                    for model in response.get("models", []):
+                    if tags:
                         model["tags"] = tags
 
-        def merge_models_lists(model_lists):
-            merged_models = {}
-
-            for idx, model_list in enumerate(model_lists):
-                if model_list is not None:
-                    for model in model_list:
-                        id = model["model"]
-                        if id not in merged_models:
-                            model["urls"] = [idx]
-                            merged_models[id] = model
-                        else:
-                            merged_models[id]["urls"].append(idx)
-
-            return list(merged_models.values())
+                    if connection_type:
+                        model["connection_type"] = connection_type
 
         models = {
-            "models": merge_models_lists(
+            "models": merge_ollama_models_lists(
                 map(
                     lambda response: response.get("models", []) if response else None,
                     responses,
@@ -384,6 +391,22 @@ async def get_all_models(request: Request, user: UserModel = None):
             )
         }
 
+        try:
+            loaded_models = await get_ollama_loaded_models(request, user=user)
+            expires_map = {
+                m["name"]: m["expires_at"]
+                for m in loaded_models["models"]
+                if "expires_at" in m
+            }
+
+            for m in models["models"]:
+                if m["name"] in expires_map:
+                    # Parse ISO8601 datetime with offset, get unix timestamp as int
+                    dt = datetime.fromisoformat(expires_map[m["name"]])
+                    m["expires_at"] = int(dt.timestamp())
+        except Exception as e:
+            log.debug(f"Failed to get loaded models: {e}")
+
     else:
         models = {"models": []}
 
@@ -464,6 +487,68 @@ async def get_ollama_tags(
     return models
 
 
+@router.get("/api/ps")
+async def get_ollama_loaded_models(request: Request, user=Depends(get_admin_user)):
+    """
+    List models that are currently loaded into Ollama memory, and which node they are loaded on.
+    """
+    if request.app.state.config.ENABLE_OLLAMA_API:
+        request_tasks = []
+        for idx, url in enumerate(request.app.state.config.OLLAMA_BASE_URLS):
+            if (str(idx) not in request.app.state.config.OLLAMA_API_CONFIGS) and (
+                url not in request.app.state.config.OLLAMA_API_CONFIGS  # Legacy support
+            ):
+                request_tasks.append(send_get_request(f"{url}/api/ps", user=user))
+            else:
+                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+                    str(idx),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        url, {}
+                    ),  # Legacy support
+                )
+
+                enable = api_config.get("enable", True)
+                key = api_config.get("key", None)
+
+                if enable:
+                    request_tasks.append(
+                        send_get_request(f"{url}/api/ps", key, user=user)
+                    )
+                else:
+                    request_tasks.append(asyncio.ensure_future(asyncio.sleep(0, None)))
+
+        responses = await asyncio.gather(*request_tasks)
+
+        for idx, response in enumerate(responses):
+            if response:
+                url = request.app.state.config.OLLAMA_BASE_URLS[idx]
+                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+                    str(idx),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        url, {}
+                    ),  # Legacy support
+                )
+
+                prefix_id = api_config.get("prefix_id", None)
+
+                for model in response.get("models", []):
+                    if prefix_id:
+                        model["model"] = f"{prefix_id}.{model['model']}"
+
+        models = {
+            "models": merge_ollama_models_lists(
+                map(
+                    lambda response: response.get("models", []) if response else None,
+                    responses,
+                )
+            )
+        }
+    else:
+        models = {"models": []}
+
+    return models
+
+
 @router.get("/api/version")
 @router.get("/api/version/{url_idx}")
 async def get_ollama_versions(request: Request, url_idx: Optional[int] = None):
@@ -537,36 +622,74 @@ async def get_ollama_versions(request: Request, url_idx: Optional[int] = None):
         return {"version": False}
 
 
-@router.get("/api/ps")
-async def get_ollama_loaded_models(request: Request, user=Depends(get_verified_user)):
-    """
-    List models that are currently loaded into Ollama memory, and which node they are loaded on.
-    """
-    if request.app.state.config.ENABLE_OLLAMA_API:
-        request_tasks = [
-            send_get_request(
-                f"{url}/api/ps",
-                request.app.state.config.OLLAMA_API_CONFIGS.get(
-                    str(idx),
-                    request.app.state.config.OLLAMA_API_CONFIGS.get(
-                        url, {}
-                    ),  # Legacy support
-                ).get("key", None),
-                user=user,
-            )
-            for idx, url in enumerate(request.app.state.config.OLLAMA_BASE_URLS)
-        ]
-        responses = await asyncio.gather(*request_tasks)
-
-        return dict(zip(request.app.state.config.OLLAMA_BASE_URLS, responses))
-    else:
-        return {}
-
-
 class ModelNameForm(BaseModel):
     name: str
 
 
+@router.post("/api/unload")
+async def unload_model(
+    request: Request,
+    form_data: ModelNameForm,
+    user=Depends(get_admin_user),
+):
+    model_name = form_data.name
+    if not model_name:
+        raise HTTPException(
+            status_code=400, detail="Missing 'name' of model to unload."
+        )
+
+    # Refresh/load models if needed, get mapping from name to URLs
+    await get_all_models(request, user=user)
+    models = request.app.state.OLLAMA_MODELS
+
+    # Canonicalize model name (if not supplied with version)
+    if ":" not in model_name:
+        model_name = f"{model_name}:latest"
+
+    if model_name not in models:
+        raise HTTPException(
+            status_code=400, detail=ERROR_MESSAGES.MODEL_NOT_FOUND(model_name)
+        )
+    url_indices = models[model_name]["urls"]
+
+    # Send unload to ALL url_indices
+    results = []
+    errors = []
+    for idx in url_indices:
+        url = request.app.state.config.OLLAMA_BASE_URLS[idx]
+        api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+            str(idx), request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+        )
+        key = get_api_key(idx, url, request.app.state.config.OLLAMA_API_CONFIGS)
+
+        prefix_id = api_config.get("prefix_id", None)
+        if prefix_id and model_name.startswith(f"{prefix_id}."):
+            model_name = model_name[len(f"{prefix_id}.") :]
+
+        payload = {"model": model_name, "keep_alive": 0, "prompt": ""}
+
+        try:
+            res = await send_post_request(
+                url=f"{url}/api/generate",
+                payload=json.dumps(payload),
+                stream=False,
+                key=key,
+                user=user,
+            )
+            results.append({"url_idx": idx, "success": True, "response": res})
+        except Exception as e:
+            log.exception(f"Failed to unload model on node {idx}: {e}")
+            errors.append({"url_idx": idx, "success": False, "error": str(e)})
+
+    if len(errors) > 0:
+        raise HTTPException(
+            status_code=500,
+            detail=f"Failed to unload model on {len(errors)} nodes: {errors}",
+        )
+
+    return {"status": True}
+
+
 @router.post("/api/pull")
 @router.post("/api/pull/{url_idx}")
 async def pull_model(
@@ -1585,7 +1708,9 @@ async def upload_model(
     if url_idx is None:
         url_idx = 0
     ollama_url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    file_path = os.path.join(UPLOAD_DIR, file.filename)
+
+    filename = os.path.basename(file.filename)
+    file_path = os.path.join(UPLOAD_DIR, filename)
     os.makedirs(UPLOAD_DIR, exist_ok=True)
 
     # --- P1: save file locally ---
@@ -1630,13 +1755,13 @@ async def upload_model(
                 os.remove(file_path)
 
                 # Create model in ollama
-                model_name, ext = os.path.splitext(file.filename)
+                model_name, ext = os.path.splitext(filename)
                 log.info(f"Created Model: {model_name}")  # DEBUG
 
                 create_payload = {
                     "model": model_name,
                     # Reference the file by its original name => the uploaded blob's digest
-                    "files": {file.filename: f"sha256:{file_hash}"},
+                    "files": {filename: f"sha256:{file_hash}"},
                 }
                 log.info(f"Model Payload: {create_payload}")  # DEBUG
 
@@ -1653,7 +1778,7 @@ async def upload_model(
                     done_msg = {
                         "done": True,
                         "blob": f"sha256:{file_hash}",
-                        "name": file.filename,
+                        "name": filename,
                         "model_created": model_name,
                     }
                     yield f"data: {json.dumps(done_msg)}\n\n"

backend/open_webui/routers/openai.py

@@ -353,21 +353,22 @@ async def get_all_models_responses(request: Request, user: UserModel) -> list:
                 ),  # Legacy support
             )
 
+            connection_type = api_config.get("connection_type", "external")
             prefix_id = api_config.get("prefix_id", None)
             tags = api_config.get("tags", [])
 
-            if prefix_id:
-                for model in (
-                    response if isinstance(response, list) else response.get("data", [])
-                ):
+            for model in (
+                response if isinstance(response, list) else response.get("data", [])
+            ):
+                if prefix_id:
                     model["id"] = f"{prefix_id}.{model['id']}"
 
-            if tags:
-                for model in (
-                    response if isinstance(response, list) else response.get("data", [])
-                ):
+                if tags:
                     model["tags"] = tags
 
+                if connection_type:
+                    model["connection_type"] = connection_type
+
     log.debug(f"get_all_models:responses() {responses}")
     return responses
 
@@ -415,6 +416,7 @@ async def get_all_models(request: Request, user: UserModel) -> dict[str, list]:
                             "name": model.get("name", model["id"]),
                             "owned_by": "openai",
                             "openai": model,
+                            "connection_type": model.get("connection_type", "external"),
                             "urlIdx": idx,
                         }
                         for model in models
@@ -461,60 +463,74 @@ async def get_models(
         url = request.app.state.config.OPENAI_API_BASE_URLS[url_idx]
         key = request.app.state.config.OPENAI_API_KEYS[url_idx]
 
+        api_config = request.app.state.config.OPENAI_API_CONFIGS.get(
+            str(url_idx),
+            request.app.state.config.OPENAI_API_CONFIGS.get(url, {}),  # Legacy support
+        )
+
         r = None
         async with aiohttp.ClientSession(
             trust_env=True,
             timeout=aiohttp.ClientTimeout(total=AIOHTTP_CLIENT_TIMEOUT_MODEL_LIST),
         ) as session:
             try:
-                async with session.get(
-                    f"{url}/models",
-                    headers={
-                        "Authorization": f"Bearer {key}",
-                        "Content-Type": "application/json",
-                        **(
-                            {
-                                "X-OpenWebUI-User-Name": user.name,
-                                "X-OpenWebUI-User-Id": user.id,
-                                "X-OpenWebUI-User-Email": user.email,
-                                "X-OpenWebUI-User-Role": user.role,
-                            }
-                            if ENABLE_FORWARD_USER_INFO_HEADERS
-                            else {}
-                        ),
-                    },
-                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
-                ) as r:
-                    if r.status != 200:
-                        # Extract response error details if available
-                        error_detail = f"HTTP Error: {r.status}"
-                        res = await r.json()
-                        if "error" in res:
-                            error_detail = f"External Error: {res['error']}"
-                        raise Exception(error_detail)
+                headers = {
+                    "Content-Type": "application/json",
+                    **(
+                        {
+                            "X-OpenWebUI-User-Name": user.name,
+                            "X-OpenWebUI-User-Id": user.id,
+                            "X-OpenWebUI-User-Email": user.email,
+                            "X-OpenWebUI-User-Role": user.role,
+                        }
+                        if ENABLE_FORWARD_USER_INFO_HEADERS
+                        else {}
+                    ),
+                }
 
-                    response_data = await r.json()
+                if api_config.get("azure", False):
+                    models = {
+                        "data": api_config.get("model_ids", []) or [],
+                        "object": "list",
+                    }
+                else:
+                    headers["Authorization"] = f"Bearer {key}"
 
-                    # Check if we're calling OpenAI API based on the URL
-                    if "api.openai.com" in url:
-                        # Filter models according to the specified conditions
-                        response_data["data"] = [
-                            model
-                            for model in response_data.get("data", [])
-                            if not any(
-                                name in model["id"]
-                                for name in [
-                                    "babbage",
-                                    "dall-e",
-                                    "davinci",
-                                    "embedding",
-                                    "tts",
-                                    "whisper",
-                                ]
-                            )
-                        ]
+                    async with session.get(
+                        f"{url}/models",
+                        headers=headers,
+                        ssl=AIOHTTP_CLIENT_SESSION_SSL,
+                    ) as r:
+                        if r.status != 200:
+                            # Extract response error details if available
+                            error_detail = f"HTTP Error: {r.status}"
+                            res = await r.json()
+                            if "error" in res:
+                                error_detail = f"External Error: {res['error']}"
+                            raise Exception(error_detail)
 
-                    models = response_data
+                        response_data = await r.json()
+
+                        # Check if we're calling OpenAI API based on the URL
+                        if "api.openai.com" in url:
+                            # Filter models according to the specified conditions
+                            response_data["data"] = [
+                                model
+                                for model in response_data.get("data", [])
+                                if not any(
+                                    name in model["id"]
+                                    for name in [
+                                        "babbage",
+                                        "dall-e",
+                                        "davinci",
+                                        "embedding",
+                                        "tts",
+                                        "whisper",
+                                    ]
+                                )
+                            ]
+
+                        models = response_data
             except aiohttp.ClientError as e:
                 # ClientError covers all aiohttp requests issues
                 log.exception(f"Client error: {str(e)}")
@@ -536,6 +552,8 @@ class ConnectionVerificationForm(BaseModel):
     url: str
     key: str
 
+    config: Optional[dict] = None
+
 
 @router.post("/verify")
 async def verify_connection(
@@ -544,39 +562,64 @@ async def verify_connection(
     url = form_data.url
     key = form_data.key
 
+    api_config = form_data.config or {}
+
     async with aiohttp.ClientSession(
         trust_env=True,
         timeout=aiohttp.ClientTimeout(total=AIOHTTP_CLIENT_TIMEOUT_MODEL_LIST),
     ) as session:
         try:
-            async with session.get(
-                f"{url}/models",
-                headers={
-                    "Authorization": f"Bearer {key}",
-                    "Content-Type": "application/json",
-                    **(
-                        {
-                            "X-OpenWebUI-User-Name": user.name,
-                            "X-OpenWebUI-User-Id": user.id,
-                            "X-OpenWebUI-User-Email": user.email,
-                            "X-OpenWebUI-User-Role": user.role,
-                        }
-                        if ENABLE_FORWARD_USER_INFO_HEADERS
-                        else {}
-                    ),
-                },
-                ssl=AIOHTTP_CLIENT_SESSION_SSL,
-            ) as r:
-                if r.status != 200:
-                    # Extract response error details if available
-                    error_detail = f"HTTP Error: {r.status}"
-                    res = await r.json()
-                    if "error" in res:
-                        error_detail = f"External Error: {res['error']}"
-                    raise Exception(error_detail)
+            headers = {
+                "Content-Type": "application/json",
+                **(
+                    {
+                        "X-OpenWebUI-User-Name": user.name,
+                        "X-OpenWebUI-User-Id": user.id,
+                        "X-OpenWebUI-User-Email": user.email,
+                        "X-OpenWebUI-User-Role": user.role,
+                    }
+                    if ENABLE_FORWARD_USER_INFO_HEADERS
+                    else {}
+                ),
+            }
 
-                response_data = await r.json()
-                return response_data
+            if api_config.get("azure", False):
+                headers["api-key"] = key
+                api_version = api_config.get("api_version", "") or "2023-03-15-preview"
+
+                async with session.get(
+                    url=f"{url}/openai/models?api-version={api_version}",
+                    headers=headers,
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
+                ) as r:
+                    if r.status != 200:
+                        # Extract response error details if available
+                        error_detail = f"HTTP Error: {r.status}"
+                        res = await r.json()
+                        if "error" in res:
+                            error_detail = f"External Error: {res['error']}"
+                        raise Exception(error_detail)
+
+                    response_data = await r.json()
+                    return response_data
+            else:
+                headers["Authorization"] = f"Bearer {key}"
+
+                async with session.get(
+                    f"{url}/models",
+                    headers=headers,
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
+                ) as r:
+                    if r.status != 200:
+                        # Extract response error details if available
+                        error_detail = f"HTTP Error: {r.status}"
+                        res = await r.json()
+                        if "error" in res:
+                            error_detail = f"External Error: {res['error']}"
+                        raise Exception(error_detail)
+
+                    response_data = await r.json()
+                    return response_data
 
         except aiohttp.ClientError as e:
             # ClientError covers all aiohttp requests issues
@@ -590,6 +633,63 @@ async def verify_connection(
             raise HTTPException(status_code=500, detail=error_detail)
 
 
+def convert_to_azure_payload(
+    url,
+    payload: dict,
+):
+    model = payload.get("model", "")
+
+    # Filter allowed parameters based on Azure OpenAI API
+    allowed_params = {
+        "messages",
+        "temperature",
+        "role",
+        "content",
+        "contentPart",
+        "contentPartImage",
+        "enhancements",
+        "dataSources",
+        "n",
+        "stream",
+        "stop",
+        "max_tokens",
+        "presence_penalty",
+        "frequency_penalty",
+        "logit_bias",
+        "user",
+        "function_call",
+        "functions",
+        "tools",
+        "tool_choice",
+        "top_p",
+        "log_probs",
+        "top_logprobs",
+        "response_format",
+        "seed",
+        "max_completion_tokens",
+    }
+
+    # Special handling for o-series models
+    if model.startswith("o") and model.endswith("-mini"):
+        # Convert max_tokens to max_completion_tokens for o-series models
+        if "max_tokens" in payload:
+            payload["max_completion_tokens"] = payload["max_tokens"]
+            del payload["max_tokens"]
+
+        # Remove temperature if not 1 for o-series models
+        if "temperature" in payload and payload["temperature"] != 1:
+            log.debug(
+                f"Removing temperature parameter for o-series model {model} as only default value (1) is supported"
+            )
+            del payload["temperature"]
+
+    # Filter out unsupported parameters
+    payload = {k: v for k, v in payload.items() if k in allowed_params}
+
+    url = f"{url}/openai/deployments/{model}"
+    return url, payload
+
+
 @router.post("/chat/completions")
 async def generate_chat_completion(
     request: Request,
@@ -690,6 +790,38 @@ async def generate_chat_completion(
             convert_logit_bias_input_to_json(payload["logit_bias"])
         )
 
+    headers = {
+        "Content-Type": "application/json",
+        **(
+            {
+                "HTTP-Referer": "https://openwebui.com/",
+                "X-Title": "Open WebUI",
+            }
+            if "openrouter.ai" in url
+            else {}
+        ),
+        **(
+            {
+                "X-OpenWebUI-User-Name": user.name,
+                "X-OpenWebUI-User-Id": user.id,
+                "X-OpenWebUI-User-Email": user.email,
+                "X-OpenWebUI-User-Role": user.role,
+            }
+            if ENABLE_FORWARD_USER_INFO_HEADERS
+            else {}
+        ),
+    }
+
+    if api_config.get("azure", False):
+        request_url, payload = convert_to_azure_payload(url, payload)
+        api_version = api_config.get("api_version", "") or "2023-03-15-preview"
+        headers["api-key"] = key
+        headers["api-version"] = api_version
+        request_url = f"{request_url}/chat/completions?api-version={api_version}"
+    else:
+        request_url = f"{url}/chat/completions"
+        headers["Authorization"] = f"Bearer {key}"
+
     payload = json.dumps(payload)
 
     r = None
@@ -704,30 +836,9 @@ async def generate_chat_completion(
 
         r = await session.request(
             method="POST",
-            url=f"{url}/chat/completions",
+            url=request_url,
             data=payload,
-            headers={
-                "Authorization": f"Bearer {key}",
-                "Content-Type": "application/json",
-                **(
-                    {
-                        "HTTP-Referer": "https://openwebui.com/",
-                        "X-Title": "Open WebUI",
-                    }
-                    if "openrouter.ai" in url
-                    else {}
-                ),
-                **(
-                    {
-                        "X-OpenWebUI-User-Name": user.name,
-                        "X-OpenWebUI-User-Id": user.id,
-                        "X-OpenWebUI-User-Email": user.email,
-                        "X-OpenWebUI-User-Role": user.role,
-                    }
-                    if ENABLE_FORWARD_USER_INFO_HEADERS
-                    else {}
-                ),
-            },
+            headers=headers,
             ssl=AIOHTTP_CLIENT_SESSION_SSL,
         )
 
@@ -783,31 +894,53 @@ async def proxy(path: str, request: Request, user=Depends(get_verified_user)):
     idx = 0
     url = request.app.state.config.OPENAI_API_BASE_URLS[idx]
     key = request.app.state.config.OPENAI_API_KEYS[idx]
+    api_config = request.app.state.config.OPENAI_API_CONFIGS.get(
+        str(idx),
+        request.app.state.config.OPENAI_API_CONFIGS.get(
+            request.app.state.config.OPENAI_API_BASE_URLS[idx], {}
+        ),  # Legacy support
+    )
 
     r = None
     session = None
     streaming = False
 
     try:
+        headers = {
+            "Content-Type": "application/json",
+            **(
+                {
+                    "X-OpenWebUI-User-Name": user.name,
+                    "X-OpenWebUI-User-Id": user.id,
+                    "X-OpenWebUI-User-Email": user.email,
+                    "X-OpenWebUI-User-Role": user.role,
+                }
+                if ENABLE_FORWARD_USER_INFO_HEADERS
+                else {}
+            ),
+        }
+
+        if api_config.get("azure", False):
+            headers["api-key"] = key
+            headers["api-version"] = (
+                api_config.get("api_version", "") or "2023-03-15-preview"
+            )
+
+            payload = json.loads(body)
+            url, payload = convert_to_azure_payload(url, payload)
+            body = json.dumps(payload).encode()
+
+            request_url = f"{url}/{path}?api-version={api_config.get('api_version', '2023-03-15-preview')}"
+        else:
+            headers["Authorization"] = f"Bearer {key}"
+            request_url = f"{url}/{path}"
+
         session = aiohttp.ClientSession(trust_env=True)
         r = await session.request(
             method=request.method,
-            url=f"{url}/{path}",
+            url=request_url,
             data=body,
-            headers={
-                "Authorization": f"Bearer {key}",
-                "Content-Type": "application/json",
-                **(
-                    {
-                        "X-OpenWebUI-User-Name": user.name,
-                        "X-OpenWebUI-User-Id": user.id,
-                        "X-OpenWebUI-User-Email": user.email,
-                        "X-OpenWebUI-User-Role": user.role,
-                    }
-                    if ENABLE_FORWARD_USER_INFO_HEADERS
-                    else {}
-                ),
-            },
+            headers=headers,
             ssl=AIOHTTP_CLIENT_SESSION_SSL,
         )
         r.raise_for_status()

backend/open_webui/routers/pipelines.py

@@ -18,7 +18,7 @@ from pydantic import BaseModel
 from starlette.responses import FileResponse
 from typing import Optional
 
-from open_webui.env import SRC_LOG_LEVELS
+from open_webui.env import SRC_LOG_LEVELS, AIOHTTP_CLIENT_SESSION_SSL
 from open_webui.config import CACHE_DIR
 from open_webui.constants import ERROR_MESSAGES
 
@@ -69,7 +69,10 @@ async def process_pipeline_inlet_filter(request, payload, user, models):
     async with aiohttp.ClientSession(trust_env=True) as session:
         for filter in sorted_filters:
             urlIdx = filter.get("urlIdx")
-            if urlIdx is None:
+
+            try:
+                urlIdx = int(urlIdx)
+            except:
                 continue
 
             url = request.app.state.config.OPENAI_API_BASE_URLS[urlIdx]
@@ -89,6 +92,7 @@ async def process_pipeline_inlet_filter(request, payload, user, models):
                     f"{url}/{filter['id']}/filter/inlet",
                     headers=headers,
                     json=request_data,
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
                 ) as response:
                     payload = await response.json()
                     response.raise_for_status()
@@ -118,7 +122,10 @@ async def process_pipeline_outlet_filter(request, payload, user, models):
     async with aiohttp.ClientSession(trust_env=True) as session:
         for filter in sorted_filters:
             urlIdx = filter.get("urlIdx")
-            if urlIdx is None:
+
+            try:
+                urlIdx = int(urlIdx)
+            except:
                 continue
 
             url = request.app.state.config.OPENAI_API_BASE_URLS[urlIdx]
@@ -138,6 +145,7 @@ async def process_pipeline_outlet_filter(request, payload, user, models):
                     f"{url}/{filter['id']}/filter/outlet",
                     headers=headers,
                     json=request_data,
+                    ssl=AIOHTTP_CLIENT_SESSION_SSL,
                 ) as response:
                     payload = await response.json()
                     response.raise_for_status()
@@ -197,8 +205,10 @@ async def upload_pipeline(
     user=Depends(get_admin_user),
 ):
     log.info(f"upload_pipeline: urlIdx={urlIdx}, filename={file.filename}")
+    filename = os.path.basename(file.filename)
+
     # Check if the uploaded file is a python file
-    if not (file.filename and file.filename.endswith(".py")):
+    if not (filename and filename.endswith(".py")):
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
             detail="Only Python (.py) files are allowed.",
@@ -206,7 +216,7 @@ async def upload_pipeline(
 
     upload_folder = f"{CACHE_DIR}/pipelines"
     os.makedirs(upload_folder, exist_ok=True)
-    file_path = os.path.join(upload_folder, file.filename)
+    file_path = os.path.join(upload_folder, filename)
 
     r = None
     try:

backend/open_webui/routers/retrieval.py

@@ -36,7 +36,7 @@ from open_webui.models.knowledge import Knowledges
 from open_webui.storage.provider import Storage
 
 
-from open_webui.retrieval.vector.connector import VECTOR_DB_CLIENT
+from open_webui.retrieval.vector.factory import VECTOR_DB_CLIENT
 
 # Document loaders
 from open_webui.retrieval.loaders.main import Loader
@@ -349,13 +349,26 @@ async def get_rag_config(request: Request, user=Depends(get_admin_user)):
         "ENABLE_RAG_HYBRID_SEARCH": request.app.state.config.ENABLE_RAG_HYBRID_SEARCH,
         "TOP_K_RERANKER": request.app.state.config.TOP_K_RERANKER,
         "RELEVANCE_THRESHOLD": request.app.state.config.RELEVANCE_THRESHOLD,
+        "HYBRID_BM25_WEIGHT": request.app.state.config.HYBRID_BM25_WEIGHT,
         # Content extraction settings
         "CONTENT_EXTRACTION_ENGINE": request.app.state.config.CONTENT_EXTRACTION_ENGINE,
         "PDF_EXTRACT_IMAGES": request.app.state.config.PDF_EXTRACT_IMAGES,
+        "DATALAB_MARKER_API_KEY": request.app.state.config.DATALAB_MARKER_API_KEY,
+        "DATALAB_MARKER_LANGS": request.app.state.config.DATALAB_MARKER_LANGS,
+        "DATALAB_MARKER_SKIP_CACHE": request.app.state.config.DATALAB_MARKER_SKIP_CACHE,
+        "DATALAB_MARKER_FORCE_OCR": request.app.state.config.DATALAB_MARKER_FORCE_OCR,
+        "DATALAB_MARKER_PAGINATE": request.app.state.config.DATALAB_MARKER_PAGINATE,
+        "DATALAB_MARKER_STRIP_EXISTING_OCR": request.app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR,
+        "DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION": request.app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION,
+        "DATALAB_MARKER_USE_LLM": request.app.state.config.DATALAB_MARKER_USE_LLM,
+        "DATALAB_MARKER_OUTPUT_FORMAT": request.app.state.config.DATALAB_MARKER_OUTPUT_FORMAT,
+        "EXTERNAL_DOCUMENT_LOADER_URL": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+        "EXTERNAL_DOCUMENT_LOADER_API_KEY": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
         "TIKA_SERVER_URL": request.app.state.config.TIKA_SERVER_URL,
         "DOCLING_SERVER_URL": request.app.state.config.DOCLING_SERVER_URL,
         "DOCLING_OCR_ENGINE": request.app.state.config.DOCLING_OCR_ENGINE,
         "DOCLING_OCR_LANG": request.app.state.config.DOCLING_OCR_LANG,
+        "DOCLING_DO_PICTURE_DESCRIPTION": request.app.state.config.DOCLING_DO_PICTURE_DESCRIPTION,
         "DOCUMENT_INTELLIGENCE_ENDPOINT": request.app.state.config.DOCUMENT_INTELLIGENCE_ENDPOINT,
         "DOCUMENT_INTELLIGENCE_KEY": request.app.state.config.DOCUMENT_INTELLIGENCE_KEY,
         "MISTRAL_OCR_API_KEY": request.app.state.config.MISTRAL_OCR_API_KEY,
@@ -371,6 +384,7 @@ async def get_rag_config(request: Request, user=Depends(get_admin_user)):
         # File upload settings
         "FILE_MAX_SIZE": request.app.state.config.FILE_MAX_SIZE,
         "FILE_MAX_COUNT": request.app.state.config.FILE_MAX_COUNT,
+        "ALLOWED_FILE_EXTENSIONS": request.app.state.config.ALLOWED_FILE_EXTENSIONS,
         # Integration settings
         "ENABLE_GOOGLE_DRIVE_INTEGRATION": request.app.state.config.ENABLE_GOOGLE_DRIVE_INTEGRATION,
         "ENABLE_ONEDRIVE_INTEGRATION": request.app.state.config.ENABLE_ONEDRIVE_INTEGRATION,
@@ -383,6 +397,7 @@ async def get_rag_config(request: Request, user=Depends(get_admin_user)):
             "WEB_SEARCH_CONCURRENT_REQUESTS": request.app.state.config.WEB_SEARCH_CONCURRENT_REQUESTS,
             "WEB_SEARCH_DOMAIN_FILTER_LIST": request.app.state.config.WEB_SEARCH_DOMAIN_FILTER_LIST,
             "BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL": request.app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL,
+            "BYPASS_WEB_SEARCH_WEB_LOADER": request.app.state.config.BYPASS_WEB_SEARCH_WEB_LOADER,
             "SEARXNG_QUERY_URL": request.app.state.config.SEARXNG_QUERY_URL,
             "YACY_QUERY_URL": request.app.state.config.YACY_QUERY_URL,
             "YACY_USERNAME": request.app.state.config.YACY_USERNAME,
@@ -435,6 +450,7 @@ class WebConfig(BaseModel):
     WEB_SEARCH_CONCURRENT_REQUESTS: Optional[int] = None
     WEB_SEARCH_DOMAIN_FILTER_LIST: Optional[List[str]] = []
     BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL: Optional[bool] = None
+    BYPASS_WEB_SEARCH_WEB_LOADER: Optional[bool] = None
     SEARXNG_QUERY_URL: Optional[str] = None
     YACY_QUERY_URL: Optional[str] = None
     YACY_USERNAME: Optional[str] = None
@@ -488,14 +504,28 @@ class ConfigForm(BaseModel):
     ENABLE_RAG_HYBRID_SEARCH: Optional[bool] = None
     TOP_K_RERANKER: Optional[int] = None
     RELEVANCE_THRESHOLD: Optional[float] = None
+    HYBRID_BM25_WEIGHT: Optional[float] = None
 
     # Content extraction settings
     CONTENT_EXTRACTION_ENGINE: Optional[str] = None
     PDF_EXTRACT_IMAGES: Optional[bool] = None
+    DATALAB_MARKER_API_KEY: Optional[str] = None
+    DATALAB_MARKER_LANGS: Optional[str] = None
+    DATALAB_MARKER_SKIP_CACHE: Optional[bool] = None
+    DATALAB_MARKER_FORCE_OCR: Optional[bool] = None
+    DATALAB_MARKER_PAGINATE: Optional[bool] = None
+    DATALAB_MARKER_STRIP_EXISTING_OCR: Optional[bool] = None
+    DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION: Optional[bool] = None
+    DATALAB_MARKER_USE_LLM: Optional[bool] = None
+    DATALAB_MARKER_OUTPUT_FORMAT: Optional[str] = None
+    EXTERNAL_DOCUMENT_LOADER_URL: Optional[str] = None
+    EXTERNAL_DOCUMENT_LOADER_API_KEY: Optional[str] = None
+
     TIKA_SERVER_URL: Optional[str] = None
     DOCLING_SERVER_URL: Optional[str] = None
     DOCLING_OCR_ENGINE: Optional[str] = None
     DOCLING_OCR_LANG: Optional[str] = None
+    DOCLING_DO_PICTURE_DESCRIPTION: Optional[bool] = None
     DOCUMENT_INTELLIGENCE_ENDPOINT: Optional[str] = None
     DOCUMENT_INTELLIGENCE_KEY: Optional[str] = None
     MISTRAL_OCR_API_KEY: Optional[str] = None
@@ -514,6 +544,7 @@ class ConfigForm(BaseModel):
     # File upload settings
     FILE_MAX_SIZE: Optional[int] = None
     FILE_MAX_COUNT: Optional[int] = None
+    ALLOWED_FILE_EXTENSIONS: Optional[List[str]] = None
 
     # Integration settings
     ENABLE_GOOGLE_DRIVE_INTEGRATION: Optional[bool] = None
@@ -569,6 +600,11 @@ async def update_rag_config(
         if form_data.RELEVANCE_THRESHOLD is not None
         else request.app.state.config.RELEVANCE_THRESHOLD
     )
+    request.app.state.config.HYBRID_BM25_WEIGHT = (
+        form_data.HYBRID_BM25_WEIGHT
+        if form_data.HYBRID_BM25_WEIGHT is not None
+        else request.app.state.config.HYBRID_BM25_WEIGHT
+    )
 
     # Content extraction settings
     request.app.state.config.CONTENT_EXTRACTION_ENGINE = (
@@ -581,6 +617,61 @@ async def update_rag_config(
         if form_data.PDF_EXTRACT_IMAGES is not None
         else request.app.state.config.PDF_EXTRACT_IMAGES
     )
+    request.app.state.config.DATALAB_MARKER_API_KEY = (
+        form_data.DATALAB_MARKER_API_KEY
+        if form_data.DATALAB_MARKER_API_KEY is not None
+        else request.app.state.config.DATALAB_MARKER_API_KEY
+    )
+    request.app.state.config.DATALAB_MARKER_LANGS = (
+        form_data.DATALAB_MARKER_LANGS
+        if form_data.DATALAB_MARKER_LANGS is not None
+        else request.app.state.config.DATALAB_MARKER_LANGS
+    )
+    request.app.state.config.DATALAB_MARKER_SKIP_CACHE = (
+        form_data.DATALAB_MARKER_SKIP_CACHE
+        if form_data.DATALAB_MARKER_SKIP_CACHE is not None
+        else request.app.state.config.DATALAB_MARKER_SKIP_CACHE
+    )
+    request.app.state.config.DATALAB_MARKER_FORCE_OCR = (
+        form_data.DATALAB_MARKER_FORCE_OCR
+        if form_data.DATALAB_MARKER_FORCE_OCR is not None
+        else request.app.state.config.DATALAB_MARKER_FORCE_OCR
+    )
+    request.app.state.config.DATALAB_MARKER_PAGINATE = (
+        form_data.DATALAB_MARKER_PAGINATE
+        if form_data.DATALAB_MARKER_PAGINATE is not None
+        else request.app.state.config.DATALAB_MARKER_PAGINATE
+    )
+    request.app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR = (
+        form_data.DATALAB_MARKER_STRIP_EXISTING_OCR
+        if form_data.DATALAB_MARKER_STRIP_EXISTING_OCR is not None
+        else request.app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR
+    )
+    request.app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION = (
+        form_data.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION
+        if form_data.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION is not None
+        else request.app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION
+    )
+    request.app.state.config.DATALAB_MARKER_OUTPUT_FORMAT = (
+        form_data.DATALAB_MARKER_OUTPUT_FORMAT
+        if form_data.DATALAB_MARKER_OUTPUT_FORMAT is not None
+        else request.app.state.config.DATALAB_MARKER_OUTPUT_FORMAT
+    )
+    request.app.state.config.DATALAB_MARKER_USE_LLM = (
+        form_data.DATALAB_MARKER_USE_LLM
+        if form_data.DATALAB_MARKER_USE_LLM is not None
+        else request.app.state.config.DATALAB_MARKER_USE_LLM
+    )
+    request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL = (
+        form_data.EXTERNAL_DOCUMENT_LOADER_URL
+        if form_data.EXTERNAL_DOCUMENT_LOADER_URL is not None
+        else request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL
+    )
+    request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY = (
+        form_data.EXTERNAL_DOCUMENT_LOADER_API_KEY
+        if form_data.EXTERNAL_DOCUMENT_LOADER_API_KEY is not None
+        else request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY
+    )
     request.app.state.config.TIKA_SERVER_URL = (
         form_data.TIKA_SERVER_URL
         if form_data.TIKA_SERVER_URL is not None
@@ -601,6 +692,13 @@ async def update_rag_config(
         if form_data.DOCLING_OCR_LANG is not None
         else request.app.state.config.DOCLING_OCR_LANG
     )
+
+    request.app.state.config.DOCLING_DO_PICTURE_DESCRIPTION = (
+        form_data.DOCLING_DO_PICTURE_DESCRIPTION
+        if form_data.DOCLING_DO_PICTURE_DESCRIPTION is not None
+        else request.app.state.config.DOCLING_DO_PICTURE_DESCRIPTION
+    )
+
     request.app.state.config.DOCUMENT_INTELLIGENCE_ENDPOINT = (
         form_data.DOCUMENT_INTELLIGENCE_ENDPOINT
         if form_data.DOCUMENT_INTELLIGENCE_ENDPOINT is not None
@@ -688,6 +786,11 @@ async def update_rag_config(
         if form_data.FILE_MAX_COUNT is not None
         else request.app.state.config.FILE_MAX_COUNT
     )
+    request.app.state.config.ALLOWED_FILE_EXTENSIONS = (
+        form_data.ALLOWED_FILE_EXTENSIONS
+        if form_data.ALLOWED_FILE_EXTENSIONS is not None
+        else request.app.state.config.ALLOWED_FILE_EXTENSIONS
+    )
 
     # Integration settings
     request.app.state.config.ENABLE_GOOGLE_DRIVE_INTEGRATION = (
@@ -720,6 +823,9 @@ async def update_rag_config(
         request.app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL = (
             form_data.web.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL
         )
+        request.app.state.config.BYPASS_WEB_SEARCH_WEB_LOADER = (
+            form_data.web.BYPASS_WEB_SEARCH_WEB_LOADER
+        )
         request.app.state.config.SEARXNG_QUERY_URL = form_data.web.SEARXNG_QUERY_URL
         request.app.state.config.YACY_QUERY_URL = form_data.web.YACY_QUERY_URL
         request.app.state.config.YACY_USERNAME = form_data.web.YACY_USERNAME
@@ -806,13 +912,26 @@ async def update_rag_config(
         "ENABLE_RAG_HYBRID_SEARCH": request.app.state.config.ENABLE_RAG_HYBRID_SEARCH,
         "TOP_K_RERANKER": request.app.state.config.TOP_K_RERANKER,
         "RELEVANCE_THRESHOLD": request.app.state.config.RELEVANCE_THRESHOLD,
+        "HYBRID_BM25_WEIGHT": request.app.state.config.HYBRID_BM25_WEIGHT,
         # Content extraction settings
         "CONTENT_EXTRACTION_ENGINE": request.app.state.config.CONTENT_EXTRACTION_ENGINE,
         "PDF_EXTRACT_IMAGES": request.app.state.config.PDF_EXTRACT_IMAGES,
+        "DATALAB_MARKER_API_KEY": request.app.state.config.DATALAB_MARKER_API_KEY,
+        "DATALAB_MARKER_LANGS": request.app.state.config.DATALAB_MARKER_LANGS,
+        "DATALAB_MARKER_SKIP_CACHE": request.app.state.config.DATALAB_MARKER_SKIP_CACHE,
+        "DATALAB_MARKER_FORCE_OCR": request.app.state.config.DATALAB_MARKER_FORCE_OCR,
+        "DATALAB_MARKER_PAGINATE": request.app.state.config.DATALAB_MARKER_PAGINATE,
+        "DATALAB_MARKER_STRIP_EXISTING_OCR": request.app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR,
+        "DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION": request.app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION,
+        "DATALAB_MARKER_USE_LLM": request.app.state.config.DATALAB_MARKER_USE_LLM,
+        "DATALAB_MARKER_OUTPUT_FORMAT": request.app.state.config.DATALAB_MARKER_OUTPUT_FORMAT,
+        "EXTERNAL_DOCUMENT_LOADER_URL": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+        "EXTERNAL_DOCUMENT_LOADER_API_KEY": request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
         "TIKA_SERVER_URL": request.app.state.config.TIKA_SERVER_URL,
         "DOCLING_SERVER_URL": request.app.state.config.DOCLING_SERVER_URL,
         "DOCLING_OCR_ENGINE": request.app.state.config.DOCLING_OCR_ENGINE,
         "DOCLING_OCR_LANG": request.app.state.config.DOCLING_OCR_LANG,
+        "DOCLING_DO_PICTURE_DESCRIPTION": request.app.state.config.DOCLING_DO_PICTURE_DESCRIPTION,
         "DOCUMENT_INTELLIGENCE_ENDPOINT": request.app.state.config.DOCUMENT_INTELLIGENCE_ENDPOINT,
         "DOCUMENT_INTELLIGENCE_KEY": request.app.state.config.DOCUMENT_INTELLIGENCE_KEY,
         "MISTRAL_OCR_API_KEY": request.app.state.config.MISTRAL_OCR_API_KEY,
@@ -828,6 +947,7 @@ async def update_rag_config(
         # File upload settings
         "FILE_MAX_SIZE": request.app.state.config.FILE_MAX_SIZE,
         "FILE_MAX_COUNT": request.app.state.config.FILE_MAX_COUNT,
+        "ALLOWED_FILE_EXTENSIONS": request.app.state.config.ALLOWED_FILE_EXTENSIONS,
         # Integration settings
         "ENABLE_GOOGLE_DRIVE_INTEGRATION": request.app.state.config.ENABLE_GOOGLE_DRIVE_INTEGRATION,
         "ENABLE_ONEDRIVE_INTEGRATION": request.app.state.config.ENABLE_ONEDRIVE_INTEGRATION,
@@ -840,6 +960,7 @@ async def update_rag_config(
             "WEB_SEARCH_CONCURRENT_REQUESTS": request.app.state.config.WEB_SEARCH_CONCURRENT_REQUESTS,
             "WEB_SEARCH_DOMAIN_FILTER_LIST": request.app.state.config.WEB_SEARCH_DOMAIN_FILTER_LIST,
             "BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL": request.app.state.config.BYPASS_WEB_SEARCH_EMBEDDING_AND_RETRIEVAL,
+            "BYPASS_WEB_SEARCH_WEB_LOADER": request.app.state.config.BYPASS_WEB_SEARCH_WEB_LOADER,
             "SEARXNG_QUERY_URL": request.app.state.config.SEARXNG_QUERY_URL,
             "YACY_QUERY_URL": request.app.state.config.YACY_QUERY_URL,
             "YACY_USERNAME": request.app.state.config.YACY_USERNAME,
@@ -1129,10 +1250,22 @@ def process_file(
                 file_path = Storage.get_file(file_path)
                 loader = Loader(
                     engine=request.app.state.config.CONTENT_EXTRACTION_ENGINE,
+                    DATALAB_MARKER_API_KEY=request.app.state.config.DATALAB_MARKER_API_KEY,
+                    DATALAB_MARKER_LANGS=request.app.state.config.DATALAB_MARKER_LANGS,
+                    DATALAB_MARKER_SKIP_CACHE=request.app.state.config.DATALAB_MARKER_SKIP_CACHE,
+                    DATALAB_MARKER_FORCE_OCR=request.app.state.config.DATALAB_MARKER_FORCE_OCR,
+                    DATALAB_MARKER_PAGINATE=request.app.state.config.DATALAB_MARKER_PAGINATE,
+                    DATALAB_MARKER_STRIP_EXISTING_OCR=request.app.state.config.DATALAB_MARKER_STRIP_EXISTING_OCR,
+                    DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION=request.app.state.config.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION,
+                    DATALAB_MARKER_USE_LLM=request.app.state.config.DATALAB_MARKER_USE_LLM,
+                    DATALAB_MARKER_OUTPUT_FORMAT=request.app.state.config.DATALAB_MARKER_OUTPUT_FORMAT,
+                    EXTERNAL_DOCUMENT_LOADER_URL=request.app.state.config.EXTERNAL_DOCUMENT_LOADER_URL,
+                    EXTERNAL_DOCUMENT_LOADER_API_KEY=request.app.state.config.EXTERNAL_DOCUMENT_LOADER_API_KEY,
                     TIKA_SERVER_URL=request.app.state.config.TIKA_SERVER_URL,
                     DOCLING_SERVER_URL=request.app.state.config.DOCLING_SERVER_URL,
                     DOCLING_OCR_ENGINE=request.app.state.config.DOCLING_OCR_ENGINE,
                     DOCLING_OCR_LANG=request.app.state.config.DOCLING_OCR_LANG,
+                    DOCLING_DO_PICTURE_DESCRIPTION=request.app.state.config.DOCLING_DO_PICTURE_DESCRIPTION,
                     PDF_EXTRACT_IMAGES=request.app.state.config.PDF_EXTRACT_IMAGES,
                     DOCUMENT_INTELLIGENCE_ENDPOINT=request.app.state.config.DOCUMENT_INTELLIGENCE_ENDPOINT,
                     DOCUMENT_INTELLIGENCE_KEY=request.app.state.config.DOCUMENT_INTELLIGENCE_KEY,
@@ -1640,13 +1773,29 @@ async def process_web_search(
         )
 
     try:
-        loader = get_web_loader(
-            urls,
-            verify_ssl=request.app.state.config.ENABLE_WEB_LOADER_SSL_VERIFICATION,
-            requests_per_second=request.app.state.config.WEB_SEARCH_CONCURRENT_REQUESTS,
-            trust_env=request.app.state.config.WEB_SEARCH_TRUST_ENV,
-        )
-        docs = await loader.aload()
+        if request.app.state.config.BYPASS_WEB_SEARCH_WEB_LOADER:
+            docs = [
+                Document(
+                    page_content=result.snippet,
+                    metadata={
+                        "source": result.link,
+                        "title": result.title,
+                        "snippet": result.snippet,
+                        "link": result.link,
+                    },
+                )
+                for result in search_results
+                if hasattr(result, "snippet")
+            ]
+        else:
+            loader = get_web_loader(
+                urls,
+                verify_ssl=request.app.state.config.ENABLE_WEB_LOADER_SSL_VERIFICATION,
+                requests_per_second=request.app.state.config.WEB_SEARCH_CONCURRENT_REQUESTS,
+                trust_env=request.app.state.config.WEB_SEARCH_TRUST_ENV,
+            )
+            docs = await loader.aload()
+
         urls = [
             doc.metadata.get("source") for doc in docs if doc.metadata.get("source")
         ]  # only keep the urls returned by the loader
@@ -1736,6 +1885,11 @@ def query_doc_handler(
                     if form_data.r
                     else request.app.state.config.RELEVANCE_THRESHOLD
                 ),
+                hybrid_bm25_weight=(
+                    form_data.hybrid_bm25_weight
+                    if form_data.hybrid_bm25_weight
+                    else request.app.state.config.HYBRID_BM25_WEIGHT
+                ),
                 user=user,
             )
         else:
@@ -1762,6 +1916,7 @@ class QueryCollectionsForm(BaseModel):
     k_reranker: Optional[int] = None
     r: Optional[float] = None
     hybrid: Optional[bool] = None
+    hybrid_bm25_weight: Optional[float] = None
 
 
 @router.post("/query/collection")
@@ -1787,6 +1942,11 @@ def query_collection_handler(
                     if form_data.r
                     else request.app.state.config.RELEVANCE_THRESHOLD
                 ),
+                hybrid_bm25_weight=(
+                    form_data.hybrid_bm25_weight
+                    if form_data.hybrid_bm25_weight
+                    else request.app.state.config.HYBRID_BM25_WEIGHT
+                ),
             )
         else:
             return query_collection(

backend/open_webui/routers/tasks.py

@@ -20,10 +20,7 @@ from open_webui.utils.auth import get_admin_user, get_verified_user
 from open_webui.constants import TASKS
 
 from open_webui.routers.pipelines import process_pipeline_inlet_filter
-from open_webui.utils.filter import (
-    get_sorted_filter_ids,
-    process_filter_functions,
-)
+
 from open_webui.utils.task import get_task_model_id
 
 from open_webui.config import (
@@ -195,15 +192,19 @@ async def generate_title(
         },
     )
 
+    max_tokens = (
+        models[task_model_id].get("info", {}).get("params", {}).get("max_tokens", 1000)
+    )
+
     payload = {
         "model": task_model_id,
         "messages": [{"role": "user", "content": content}],
         "stream": False,
         **(
-            {"max_tokens": 1000}
+            {"max_tokens": max_tokens}
             if models[task_model_id].get("owned_by") == "ollama"
             else {
-                "max_completion_tokens": 1000,
+                "max_completion_tokens": max_tokens,
             }
         ),
         "metadata": {

backend/open_webui/routers/tools.py

@@ -2,6 +2,9 @@ import logging
 from pathlib import Path
 from typing import Optional
 import time
+import re
+import aiohttp
+from pydantic import BaseModel, HttpUrl
 
 from open_webui.models.tools import (
     ToolForm,
@@ -21,6 +24,7 @@ from open_webui.env import SRC_LOG_LEVELS
 
 from open_webui.utils.tools import get_tool_servers_data
 
+
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MAIN"])
 
@@ -51,11 +55,11 @@ async def get_tools(request: Request, user=Depends(get_verified_user)):
                 **{
                     "id": f"server:{server['idx']}",
                     "user_id": f"server:{server['idx']}",
-                    "name": server["openapi"]
+                    "name": server.get("openapi", {})
                     .get("info", {})
                     .get("title", "Tool Server"),
                     "meta": {
-                        "description": server["openapi"]
+                        "description": server.get("openapi", {})
                         .get("info", {})
                         .get("description", ""),
                     },
@@ -95,6 +99,81 @@ async def get_tool_list(user=Depends(get_verified_user)):
     return tools
 
 
+############################
+# LoadFunctionFromLink
+############################
+
+
+class LoadUrlForm(BaseModel):
+    url: HttpUrl
+
+
+def github_url_to_raw_url(url: str) -> str:
+    # Handle 'tree' (folder) URLs (add main.py at the end)
+    m1 = re.match(r"https://github\.com/([^/]+)/([^/]+)/tree/([^/]+)/(.*)", url)
+    if m1:
+        org, repo, branch, path = m1.groups()
+        return f"https://raw.githubusercontent.com/{org}/{repo}/refs/heads/{branch}/{path.rstrip('/')}/main.py"
+
+    # Handle 'blob' (file) URLs
+    m2 = re.match(r"https://github\.com/([^/]+)/([^/]+)/blob/([^/]+)/(.*)", url)
+    if m2:
+        org, repo, branch, path = m2.groups()
+        return (
+            f"https://raw.githubusercontent.com/{org}/{repo}/refs/heads/{branch}/{path}"
+        )
+
+    # No match; return as-is
+    return url
+
+
+@router.post("/load/url", response_model=Optional[dict])
+async def load_tool_from_url(
+    request: Request, form_data: LoadUrlForm, user=Depends(get_admin_user)
+):
+    # NOTE: This is NOT a SSRF vulnerability:
+    # This endpoint is admin-only (see get_admin_user), meant for *trusted* internal use,
+    # and does NOT accept untrusted user input. Access is enforced by authentication.
+
+    url = str(form_data.url)
+    if not url:
+        raise HTTPException(status_code=400, detail="Please enter a valid URL")
+
+    url = github_url_to_raw_url(url)
+    url_parts = url.rstrip("/").split("/")
+
+    file_name = url_parts[-1]
+    tool_name = (
+        file_name[:-3]
+        if (
+            file_name.endswith(".py")
+            and (not file_name.startswith(("main.py", "index.py", "__init__.py")))
+        )
+        else url_parts[-2] if len(url_parts) > 1 else "function"
+    )
+
+    try:
+        async with aiohttp.ClientSession() as session:
+            async with session.get(
+                url, headers={"Content-Type": "application/json"}
+            ) as resp:
+                if resp.status != 200:
+                    raise HTTPException(
+                        status_code=resp.status, detail="Failed to fetch the tool"
+                    )
+                data = await resp.text()
+                if not data:
+                    raise HTTPException(
+                        status_code=400, detail="No data received from the URL"
+                    )
+        return {
+            "name": tool_name,
+            "content": data,
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error importing tool: {e}")
+
+
 ############################
 # ExportTools
 ############################

backend/open_webui/static/assets/pdf-style.css

@@ -269,11 +269,6 @@ tbody + tbody {
 	margin-bottom: 0;
 }
 
-/* Add a rule to reset margin-bottom for <p> not followed by <ul> */
-.markdown-section p + ul {
-	margin-top: 0;
-}
-
 /* List item styles */
 .markdown-section li {
 	padding: 2px;

backend/open_webui/storage/provider.py

@@ -2,6 +2,7 @@ import os
 import shutil
 import json
 import logging
+import re
 from abc import ABC, abstractmethod
 from typing import BinaryIO, Tuple, Dict
 
@@ -136,6 +137,11 @@ class S3StorageProvider(StorageProvider):
         self.bucket_name = S3_BUCKET_NAME
         self.key_prefix = S3_KEY_PREFIX if S3_KEY_PREFIX else ""
 
+    @staticmethod
+    def sanitize_tag_value(s: str) -> str:
+        """Only include S3 allowed characters."""
+        return re.sub(r"[^a-zA-Z0-9 äöüÄÖÜß\+\-=\._:/@]", "", s)
+
     def upload_file(
         self, file: BinaryIO, filename: str, tags: Dict[str, str]
     ) -> Tuple[bytes, str]:
@@ -145,7 +151,15 @@ class S3StorageProvider(StorageProvider):
         try:
             self.s3_client.upload_file(file_path, self.bucket_name, s3_key)
             if S3_ENABLE_TAGGING and tags:
-                tagging = {"TagSet": [{"Key": k, "Value": v} for k, v in tags.items()]}
+                sanitized_tags = {
+                    self.sanitize_tag_value(k): self.sanitize_tag_value(v)
+                    for k, v in tags.items()
+                }
+                tagging = {
+                    "TagSet": [
+                        {"Key": k, "Value": v} for k, v in sanitized_tags.items()
+                    ]
+                }
                 self.s3_client.put_object_tagging(
                     Bucket=self.bucket_name,
                     Key=s3_key,

backend/open_webui/utils/auth.py

@@ -13,6 +13,8 @@ import pytz
 from pytz import UTC
 from typing import Optional, Union, List, Dict
 
+from opentelemetry import trace
+
 from open_webui.models.users import Users
 
 from open_webui.constants import ERROR_MESSAGES
@@ -194,7 +196,17 @@ def get_current_user(
                     status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.API_KEY_NOT_ALLOWED
                 )
 
-        return get_current_user_by_api_key(token)
+        user = get_current_user_by_api_key(token)
+
+        # Add user info to current span
+        current_span = trace.get_current_span()
+        if current_span:
+            current_span.set_attribute("client.user.id", user.id)
+            current_span.set_attribute("client.user.email", user.email)
+            current_span.set_attribute("client.user.role", user.role)
+            current_span.set_attribute("client.auth.type", "api_key")
+
+        return user
 
     # auth by jwt token
     try:
@@ -213,6 +225,14 @@ def get_current_user(
                 detail=ERROR_MESSAGES.INVALID_TOKEN,
             )
         else:
+            # Add user info to current span
+            current_span = trace.get_current_span()
+            if current_span:
+                current_span.set_attribute("client.user.id", user.id)
+                current_span.set_attribute("client.user.email", user.email)
+                current_span.set_attribute("client.user.role", user.role)
+                current_span.set_attribute("client.auth.type", "jwt")
+
             # Refresh the user's last active timestamp asynchronously
             # to prevent blocking the request
             if background_tasks:
@@ -234,6 +254,14 @@ def get_current_user_by_api_key(api_key: str):
             detail=ERROR_MESSAGES.INVALID_TOKEN,
         )
     else:
+        # Add user info to current span
+        current_span = trace.get_current_span()
+        if current_span:
+            current_span.set_attribute("client.user.id", user.id)
+            current_span.set_attribute("client.user.email", user.email)
+            current_span.set_attribute("client.user.role", user.role)
+            current_span.set_attribute("client.auth.type", "api_key")
+
         Users.update_user_last_active_by_id(user.id)
 
     return user

backend/open_webui/utils/chat.py

@@ -40,7 +40,10 @@ from open_webui.models.functions import Functions
 from open_webui.models.models import Models
 
 
-from open_webui.utils.plugin import load_function_module_by_id
+from open_webui.utils.plugin import (
+    load_function_module_by_id,
+    get_function_module_from_cache,
+)
 from open_webui.utils.models import get_all_models, check_model_access
 from open_webui.utils.payload import convert_payload_openai_to_ollama
 from open_webui.utils.response import (
@@ -309,6 +312,7 @@ async def chat_completed(request: Request, form_data: dict, user: Any):
     metadata = {
         "chat_id": data["chat_id"],
         "message_id": data["id"],
+        "filter_ids": data.get("filter_ids", []),
         "session_id": data["session_id"],
         "user_id": user.id,
     }
@@ -330,7 +334,9 @@ async def chat_completed(request: Request, form_data: dict, user: Any):
     try:
         filter_functions = [
             Functions.get_function_by_id(filter_id)
-            for filter_id in get_sorted_filter_ids(model)
+            for filter_id in get_sorted_filter_ids(
+                request, model, metadata.get("filter_ids", [])
+            )
         ]
 
         result, _ = await process_filter_functions(
@@ -389,11 +395,7 @@ async def chat_action(request: Request, action_id: str, form_data: dict, user: A
         }
     )
 
-    if action_id in request.app.state.FUNCTIONS:
-        function_module = request.app.state.FUNCTIONS[action_id]
-    else:
-        function_module, _, _ = load_function_module_by_id(action_id)
-        request.app.state.FUNCTIONS[action_id] = function_module
+    function_module, _, _ = get_function_module_from_cache(request, action_id)
 
     if hasattr(function_module, "valves") and hasattr(function_module, "Valves"):
         valves = Functions.get_function_valves_by_id(action_id)

backend/open_webui/utils/filter.py

@@ -1,7 +1,10 @@
 import inspect
 import logging
 
-from open_webui.utils.plugin import load_function_module_by_id
+from open_webui.utils.plugin import (
+    load_function_module_by_id,
+    get_function_module_from_cache,
+)
 from open_webui.models.functions import Functions
 from open_webui.env import SRC_LOG_LEVELS
 
@@ -9,7 +12,17 @@ log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MAIN"])
 
 
-def get_sorted_filter_ids(model: dict):
+def get_function_module(request, function_id, load_from_db=True):
+    """
+    Get the function module by its ID.
+    """
+    function_module, _, _ = get_function_module_from_cache(
+        request, function_id, load_from_db
+    )
+    return function_module
+
+
+def get_sorted_filter_ids(request, model: dict, enabled_filter_ids: list = None):
     def get_priority(function_id):
         function = Functions.get_function_by_id(function_id)
         if function is not None:
@@ -21,14 +34,26 @@ def get_sorted_filter_ids(model: dict):
     if "info" in model and "meta" in model["info"]:
         filter_ids.extend(model["info"]["meta"].get("filterIds", []))
         filter_ids = list(set(filter_ids))
-
-    enabled_filter_ids = [
+    active_filter_ids = [
         function.id
         for function in Functions.get_functions_by_type("filter", active_only=True)
     ]
 
-    filter_ids = [fid for fid in filter_ids if fid in enabled_filter_ids]
+    def get_active_status(filter_id):
+        function_module = get_function_module(request, filter_id)
+
+        if getattr(function_module, "toggle", None):
+            return filter_id in (enabled_filter_ids or [])
+
+        return True
+
+    active_filter_ids = [
+        filter_id for filter_id in active_filter_ids if get_active_status(filter_id)
+    ]
+
+    filter_ids = [fid for fid in filter_ids if fid in active_filter_ids]
     filter_ids.sort(key=get_priority)
+
     return filter_ids
 
 
@@ -43,12 +68,9 @@ async def process_filter_functions(
         if not filter:
             continue
 
-        if filter_id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[filter_id]
-        else:
-            function_module, _, _ = load_function_module_by_id(filter_id)
-            request.app.state.FUNCTIONS[filter_id] = function_module
-
+        function_module = get_function_module(
+            request, filter_id, load_from_db=(filter_type != "stream")
+        )
         # Prepare handler function
         handler = getattr(function_module, filter_type, None)
         if not handler:

backend/open_webui/utils/middleware.py

@@ -41,6 +41,7 @@ from open_webui.routers.pipelines import (
     process_pipeline_inlet_filter,
     process_pipeline_outlet_filter,
 )
+from open_webui.routers.memories import query_memory, QueryMemoryForm
 
 from open_webui.utils.webhook import post_webhook
 
@@ -251,7 +252,12 @@ async def chat_completion_tools_handler(
                                     "name": (f"TOOL:{tool_name}"),
                                 },
                                 "document": [tool_result],
-                                "metadata": [{"source": (f"TOOL:{tool_name}")}],
+                                "metadata": [
+                                    {
+                                        "source": (f"TOOL:{tool_name}"),
+                                        "parameters": tool_function_params,
+                                    }
+                                ],
                             }
                         )
                     else:
@@ -290,6 +296,38 @@ async def chat_completion_tools_handler(
     return body, {"sources": sources}
 
 
+async def chat_memory_handler(
+    request: Request, form_data: dict, extra_params: dict, user
+):
+    results = await query_memory(
+        request,
+        QueryMemoryForm(
+            **{"content": get_last_user_message(form_data["messages"]), "k": 3}
+        ),
+        user,
+    )
+
+    user_context = ""
+    if results and hasattr(results, "documents"):
+        if results.documents and len(results.documents) > 0:
+            for doc_idx, doc in enumerate(results.documents[0]):
+                created_at_date = "Unknown Date"
+
+                if results.metadatas[0][doc_idx].get("created_at"):
+                    created_at_timestamp = results.metadatas[0][doc_idx]["created_at"]
+                    created_at_date = time.strftime(
+                        "%Y-%m-%d", time.localtime(created_at_timestamp)
+                    )
+
+                user_context += f"{doc_idx + 1}. [{created_at_date}] {doc}\n"
+
+    form_data["messages"] = add_or_update_system_message(
+        f"User Context:\n{user_context}\n", form_data["messages"], append=True
+    )
+
+    return form_data
+
+
 async def chat_web_search_handler(
     request: Request, form_data: dict, extra_params: dict, user
 ):
@@ -340,6 +378,11 @@ async def chat_web_search_handler(
         log.exception(e)
         queries = [user_message]
 
+    # Check if generated queries are empty
+    if len(queries) == 1 and queries[0].strip() == "":
+        queries = [user_message]
+
+    # Check if queries are not found
     if len(queries) == 0:
         await event_emitter(
             {
@@ -384,6 +427,7 @@ async def chat_web_search_handler(
                             "name": ", ".join(queries),
                             "type": "web_search",
                             "urls": results["filenames"],
+                            "queries": queries,
                         }
                     )
             elif results.get("docs"):
@@ -395,6 +439,7 @@ async def chat_web_search_handler(
                         "name": ", ".join(queries),
                         "type": "web_search",
                         "urls": results["filenames"],
+                        "queries": queries,
                     }
                 )
 
@@ -598,6 +643,7 @@ async def chat_completion_files_handler(
                         reranking_function=request.app.state.rf,
                         k_reranker=request.app.state.config.TOP_K_RERANKER,
                         r=request.app.state.config.RELEVANCE_THRESHOLD,
+                        hybrid_bm25_weight=request.app.state.config.HYBRID_BM25_WEIGHT,
                         hybrid_search=request.app.state.config.ENABLE_RAG_HYBRID_SEARCH,
                         full_context=request.app.state.config.RAG_FULL_CONTEXT,
                     ),
@@ -612,6 +658,12 @@ async def chat_completion_files_handler(
 
 def apply_params_to_form_data(form_data, model):
     params = form_data.pop("params", {})
+    custom_params = params.pop("custom_params", {})
+
+    if custom_params:
+        # If custom_params are provided, merge them into params
+        params = deep_update(params, custom_params)
+
     if model.get("ollama"):
         form_data["options"] = params
 
@@ -621,29 +673,10 @@ def apply_params_to_form_data(form_data, model):
         if "keep_alive" in params:
             form_data["keep_alive"] = params["keep_alive"]
     else:
-        if "seed" in params and params["seed"] is not None:
-            form_data["seed"] = params["seed"]
-
-        if "stop" in params and params["stop"] is not None:
-            form_data["stop"] = params["stop"]
-
-        if "temperature" in params and params["temperature"] is not None:
-            form_data["temperature"] = params["temperature"]
-
-        if "max_tokens" in params and params["max_tokens"] is not None:
-            form_data["max_tokens"] = params["max_tokens"]
-
-        if "top_p" in params and params["top_p"] is not None:
-            form_data["top_p"] = params["top_p"]
-
-        if "frequency_penalty" in params and params["frequency_penalty"] is not None:
-            form_data["frequency_penalty"] = params["frequency_penalty"]
-
-        if "presence_penalty" in params and params["presence_penalty"] is not None:
-            form_data["presence_penalty"] = params["presence_penalty"]
-
-        if "reasoning_effort" in params and params["reasoning_effort"] is not None:
-            form_data["reasoning_effort"] = params["reasoning_effort"]
+        if isinstance(params, dict):
+            for key, value in params.items():
+                if value is not None:
+                    form_data[key] = value
 
         if "logit_bias" in params and params["logit_bias"] is not None:
             try:
@@ -651,13 +684,12 @@ def apply_params_to_form_data(form_data, model):
                     convert_logit_bias_input_to_json(params["logit_bias"])
                 )
             except Exception as e:
-                print(f"Error parsing logit_bias: {e}")
+                log.exception(f"Error parsing logit_bias: {e}")
 
     return form_data
 
 
 async def process_chat_payload(request, form_data, user, metadata, model):
-
     form_data = apply_params_to_form_data(form_data, model)
     log.debug(f"form_data: {form_data}")
 
@@ -749,9 +781,12 @@ async def process_chat_payload(request, form_data, user, metadata, model):
         raise e
 
     try:
+
         filter_functions = [
             Functions.get_function_by_id(filter_id)
-            for filter_id in get_sorted_filter_ids(model)
+            for filter_id in get_sorted_filter_ids(
+                request, model, metadata.get("filter_ids", [])
+            )
         ]
 
         form_data, flags = await process_filter_functions(
@@ -766,6 +801,11 @@ async def process_chat_payload(request, form_data, user, metadata, model):
 
     features = form_data.pop("features", None)
     if features:
+        if "memory" in features and features["memory"]:
+            form_data = await chat_memory_handler(
+                request, form_data, extra_params, user
+            )
+
         if "web_search" in features and features["web_search"]:
             form_data = await chat_web_search_handler(
                 request, form_data, extra_params, user
@@ -868,6 +908,7 @@ async def process_chat_payload(request, form_data, user, metadata, model):
                 for doc_context, doc_meta in zip(
                     source["document"], source["metadata"]
                 ):
+                    source_name = source.get("source", {}).get("name", None)
                     citation_id = (
                         doc_meta.get("source", None)
                         or source.get("source", {}).get("id", None)
@@ -875,7 +916,11 @@ async def process_chat_payload(request, form_data, user, metadata, model):
                     )
                     if citation_id not in citation_idx:
                         citation_idx[citation_id] = len(citation_idx) + 1
-                    context_string += f'<source id="{citation_idx[citation_id]}">{doc_context}</source>\n'
+                    context_string += (
+                        f'<source id="{citation_idx[citation_id]}"'
+                        + (f' name="{source_name}"' if source_name else "")
+                        + f">{doc_context}</source>\n"
+                    )
 
         context_string = context_string.strip()
         prompt = get_last_user_message(form_data["messages"])
@@ -942,21 +987,38 @@ async def process_chat_response(
         message = message_map.get(metadata["message_id"]) if message_map else None
 
         if message:
-            messages = get_message_list(message_map, message.get("id"))
+            message_list = get_message_list(message_map, metadata["message_id"])
 
-            # Remove reasoning details and files from the messages.
+            # Remove details tags and files from the messages.
             # as get_message_list creates a new list, it does not affect
             # the original messages outside of this handler
-            for message in messages:
-                message["content"] = re.sub(
-                    r"<details\s+type=\"reasoning\"[^>]*>.*?<\/details>",
-                    "",
-                    message["content"],
-                    flags=re.S,
-                ).strip()
 
-                if message.get("files"):
-                    message["files"] = []
+            messages = []
+            for message in message_list:
+                content = message.get("content", "")
+                if isinstance(content, list):
+                    for item in content:
+                        if item.get("type") == "text":
+                            content = item["text"]
+                            break
+
+                if isinstance(content, str):
+                    content = re.sub(
+                        r"<details\b[^>]*>.*?<\/details>|!\[.*?\]\(.*?\)",
+                        "",
+                        content,
+                        flags=re.S | re.I,
+                    ).strip()
+
+                messages.append(
+                    {
+                        **message,
+                        "role": message.get(
+                            "role", "assistant"
+                        ),  # Safe fallback for missing role
+                        "content": content,
+                    }
+                )
 
             if tasks and messages:
                 if TASKS.TITLE_GENERATION in tasks:
@@ -1121,6 +1183,7 @@ async def process_chat_response(
                         metadata["chat_id"],
                         metadata["message_id"],
                         {
+                            "role": "assistant",
                             "content": content,
                         },
                     )
@@ -1143,8 +1206,34 @@ async def process_chat_response(
 
                     await background_tasks_handler()
 
+            if events and isinstance(events, list) and isinstance(response, dict):
+                extra_response = {}
+                for event in events:
+                    if isinstance(event, dict):
+                        extra_response.update(event)
+                    else:
+                        extra_response[event] = True
+
+                response = {
+                    **extra_response,
+                    **response,
+                }
+
             return response
         else:
+            if events and isinstance(events, list) and isinstance(response, dict):
+                extra_response = {}
+                for event in events:
+                    if isinstance(event, dict):
+                        extra_response.update(event)
+                    else:
+                        extra_response[event] = True
+
+                response = {
+                    **extra_response,
+                    **response,
+                }
+
             return response
 
     # Non standard response
@@ -1169,7 +1258,9 @@ async def process_chat_response(
     }
     filter_functions = [
         Functions.get_function_by_id(filter_id)
-        for filter_id in get_sorted_filter_ids(model)
+        for filter_id in get_sorted_filter_ids(
+            request, model, metadata.get("filter_ids", [])
+        )
     ]
 
     # Streaming response

backend/open_webui/utils/misc.py

@@ -34,11 +34,15 @@ def get_message_list(messages, message_id):
     :return: List of ordered messages starting from the root to the given message
     """
 
+    # Handle case where messages is None
+    if not messages:
+        return []  # Return empty list instead of None to prevent iteration errors
+
     # Find the message by its id
     current_message = messages.get(message_id)
 
     if not current_message:
-        return None
+        return []  # Return empty list instead of None to prevent iteration errors
 
     # Reconstruct the chain by following the parentId links
     message_list = []
@@ -47,7 +51,7 @@ def get_message_list(messages, message_id):
         message_list.insert(
             0, current_message
         )  # Insert the message at the beginning of the list
-        parent_id = current_message["parentId"]
+        parent_id = current_message.get("parentId")  # Use .get() for safety
         current_message = messages.get(parent_id) if parent_id else None
 
     return message_list
@@ -70,12 +74,12 @@ def get_last_user_message_item(messages: list[dict]) -> Optional[dict]:
 
 
 def get_content_from_message(message: dict) -> Optional[str]:
-    if isinstance(message["content"], list):
+    if isinstance(message.get("content"), list):
         for item in message["content"]:
             if item["type"] == "text":
                 return item["text"]
     else:
-        return message["content"]
+        return message.get("content")
     return None
 
 
@@ -130,7 +134,9 @@ def prepend_to_first_user_message_content(
     return messages
 
 
-def add_or_update_system_message(content: str, messages: list[dict]):
+def add_or_update_system_message(
+    content: str, messages: list[dict], append: bool = False
+):
     """
     Adds a new system message at the beginning of the messages list
     or updates the existing system message at the beginning.
@@ -141,7 +147,10 @@ def add_or_update_system_message(content: str, messages: list[dict]):
     """
 
     if messages and messages[0].get("role") == "system":
-        messages[0]["content"] = f"{content}\n{messages[0]['content']}"
+        if append:
+            messages[0]["content"] = f"{messages[0]['content']}\n{content}"
+        else:
+            messages[0]["content"] = f"{content}\n{messages[0]['content']}"
     else:
         # Insert at the beginning
         messages.insert(0, {"role": "system", "content": content})

backend/open_webui/utils/models.py

@@ -1,5 +1,6 @@
 import time
 import logging
+import asyncio
 import sys
 
 from aiocache import cached
@@ -13,7 +14,10 @@ from open_webui.models.functions import Functions
 from open_webui.models.models import Models
 
 
-from open_webui.utils.plugin import load_function_module_by_id
+from open_webui.utils.plugin import (
+    load_function_module_by_id,
+    get_function_module_from_cache,
+)
 from open_webui.utils.access_control import has_access
 
 
@@ -30,34 +34,46 @@ log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MAIN"])
 
 
+async def fetch_ollama_models(request: Request, user: UserModel = None):
+    raw_ollama_models = await ollama.get_all_models(request, user=user)
+    return [
+        {
+            "id": model["model"],
+            "name": model["name"],
+            "object": "model",
+            "created": int(time.time()),
+            "owned_by": "ollama",
+            "ollama": model,
+            "connection_type": model.get("connection_type", "local"),
+            "tags": model.get("tags", []),
+        }
+        for model in raw_ollama_models["models"]
+    ]
+
+
+async def fetch_openai_models(request: Request, user: UserModel = None):
+    openai_response = await openai.get_all_models(request, user=user)
+    return openai_response["data"]
+
+
 async def get_all_base_models(request: Request, user: UserModel = None):
-    function_models = []
-    openai_models = []
-    ollama_models = []
+    openai_task = (
+        fetch_openai_models(request, user)
+        if request.app.state.config.ENABLE_OPENAI_API
+        else asyncio.sleep(0, result=[])
+    )
+    ollama_task = (
+        fetch_ollama_models(request, user)
+        if request.app.state.config.ENABLE_OLLAMA_API
+        else asyncio.sleep(0, result=[])
+    )
+    function_task = get_function_models(request)
 
-    if request.app.state.config.ENABLE_OPENAI_API:
-        openai_models = await openai.get_all_models(request, user=user)
-        openai_models = openai_models["data"]
+    openai_models, ollama_models, function_models = await asyncio.gather(
+        openai_task, ollama_task, function_task
+    )
 
-    if request.app.state.config.ENABLE_OLLAMA_API:
-        ollama_models = await ollama.get_all_models(request, user=user)
-        ollama_models = [
-            {
-                "id": model["model"],
-                "name": model["name"],
-                "object": "model",
-                "created": int(time.time()),
-                "owned_by": "ollama",
-                "ollama": model,
-                "tags": model.get("tags", []),
-            }
-            for model in ollama_models["models"]
-        ]
-
-    function_models = await get_function_models(request)
-    models = function_models + openai_models + ollama_models
-
-    return models
+    return function_models + openai_models + ollama_models
 
 
 async def get_all_models(request, user: UserModel = None):
@@ -110,6 +126,14 @@ async def get_all_models(request, user: UserModel = None):
         for function in Functions.get_functions_by_type("action", active_only=True)
     ]
 
+    global_filter_ids = [
+        function.id for function in Functions.get_global_filter_functions()
+    ]
+    enabled_filter_ids = [
+        function.id
+        for function in Functions.get_functions_by_type("filter", active_only=True)
+    ]
+
     custom_models = Models.get_all_models()
     for custom_model in custom_models:
         if custom_model.base_model_id is None:
@@ -125,13 +149,20 @@ async def get_all_models(request, user: UserModel = None):
                         model["name"] = custom_model.name
                         model["info"] = custom_model.model_dump()
 
+                        # Set action_ids and filter_ids
                         action_ids = []
+                        filter_ids = []
+
                         if "info" in model and "meta" in model["info"]:
                             action_ids.extend(
                                 model["info"]["meta"].get("actionIds", [])
                             )
+                            filter_ids.extend(
+                                model["info"]["meta"].get("filterIds", [])
+                            )
 
                         model["action_ids"] = action_ids
+                        model["filter_ids"] = filter_ids
                     else:
                         models.remove(model)
 
@@ -140,7 +171,9 @@ async def get_all_models(request, user: UserModel = None):
         ):
             owned_by = "openai"
             pipe = None
+
             action_ids = []
+            filter_ids = []
 
             for model in models:
                 if (
@@ -154,9 +187,13 @@ async def get_all_models(request, user: UserModel = None):
 
             if custom_model.meta:
                 meta = custom_model.meta.model_dump()
+
                 if "actionIds" in meta:
                     action_ids.extend(meta["actionIds"])
 
+                if "filterIds" in meta:
+                    filter_ids.extend(meta["filterIds"])
+
             models.append(
                 {
                     "id": f"{custom_model.id}",
@@ -168,6 +205,7 @@ async def get_all_models(request, user: UserModel = None):
                     "preset": True,
                     **({"pipe": pipe} if pipe is not None else {}),
                     "action_ids": action_ids,
+                    "filter_ids": filter_ids,
                 }
             )
 
@@ -181,8 +219,11 @@ async def get_all_models(request, user: UserModel = None):
                     "id": f"{function.id}.{action['id']}",
                     "name": action.get("name", f"{function.name} ({action['id']})"),
                     "description": function.meta.description,
-                    "icon_url": action.get(
-                        "icon_url", function.meta.manifest.get("icon_url", None)
+                    "icon": action.get(
+                        "icon_url",
+                        function.meta.manifest.get("icon_url", None)
+                        or getattr(module, "icon_url", None)
+                        or getattr(module, "icon", None),
                     ),
                 }
                 for action in actions
@@ -193,16 +234,27 @@ async def get_all_models(request, user: UserModel = None):
                     "id": function.id,
                     "name": function.name,
                     "description": function.meta.description,
-                    "icon_url": function.meta.manifest.get("icon_url", None),
+                    "icon": function.meta.manifest.get("icon_url", None)
+                    or getattr(module, "icon_url", None)
+                    or getattr(module, "icon", None),
                 }
             ]
 
+    # Process filter_ids to get the filters
+    def get_filter_items_from_module(function, module):
+        return [
+            {
+                "id": function.id,
+                "name": function.name,
+                "description": function.meta.description,
+                "icon": function.meta.manifest.get("icon_url", None)
+                or getattr(module, "icon_url", None)
+                or getattr(module, "icon", None),
+            }
+        ]
+
     def get_function_module_by_id(function_id):
-        if function_id in request.app.state.FUNCTIONS:
-            function_module = request.app.state.FUNCTIONS[function_id]
-        else:
-            function_module, _, _ = load_function_module_by_id(function_id)
-            request.app.state.FUNCTIONS[function_id] = function_module
+        function_module, _, _ = get_function_module_from_cache(request, function_id)
         return function_module
 
     for model in models:
@@ -211,6 +263,11 @@ async def get_all_models(request, user: UserModel = None):
             for action_id in list(set(model.pop("action_ids", []) + global_action_ids))
             if action_id in enabled_action_ids
         ]
+        filter_ids = [
+            filter_id
+            for filter_id in list(set(model.pop("filter_ids", []) + global_filter_ids))
+            if filter_id in enabled_filter_ids
+        ]
 
         model["actions"] = []
         for action_id in action_ids:
@@ -222,6 +279,20 @@ async def get_all_models(request, user: UserModel = None):
             model["actions"].extend(
                 get_action_items_from_module(action_function, function_module)
             )
+
+        model["filters"] = []
+        for filter_id in filter_ids:
+            filter_function = Functions.get_function_by_id(filter_id)
+            if filter_function is None:
+                raise Exception(f"Filter not found: {filter_id}")
+
+            function_module = get_function_module_by_id(filter_id)
+
+            if getattr(function_module, "toggle", None):
+                model["filters"].extend(
+                    get_filter_items_from_module(filter_function, function_module)
+                )
+
     log.debug(f"get_all_models() returned {len(models)} models")
 
     request.app.state.MODELS = {model["id"]: model for model in models}

backend/open_webui/utils/oauth.py

@@ -41,6 +41,7 @@ from open_webui.config import (
 )
 from open_webui.constants import ERROR_MESSAGES, WEBHOOK_MESSAGES
 from open_webui.env import (
+    AIOHTTP_CLIENT_SESSION_SSL,
     WEBUI_NAME,
     WEBUI_AUTH_COOKIE_SAME_SITE,
     WEBUI_AUTH_COOKIE_SECURE,
@@ -305,8 +306,10 @@ class OAuthManager:
                 get_kwargs["headers"] = {
                     "Authorization": f"Bearer {access_token}",
                 }
-            async with aiohttp.ClientSession() as session:
-                async with session.get(picture_url, **get_kwargs) as resp:
+            async with aiohttp.ClientSession(trust_env=True) as session:
+                async with session.get(
+                    picture_url, **get_kwargs, ssl=AIOHTTP_CLIENT_SESSION_SSL
+                ) as resp:
                     if resp.ok:
                         picture = await resp.read()
                         base64_encoded_picture = base64.b64encode(picture).decode(
@@ -371,7 +374,9 @@ class OAuthManager:
                     headers = {"Authorization": f"Bearer {access_token}"}
                     async with aiohttp.ClientSession(trust_env=True) as session:
                         async with session.get(
-                            "https://api.github.com/user/emails", headers=headers
+                            "https://api.github.com/user/emails",
+                            headers=headers,
+                            ssl=AIOHTTP_CLIENT_SESSION_SSL,
                         ) as resp:
                             if resp.ok:
                                 emails = await resp.json()
@@ -531,5 +536,10 @@ class OAuthManager:
                 secure=WEBUI_AUTH_COOKIE_SECURE,
             )
         # Redirect back to the frontend with the JWT token
-        redirect_url = f"{request.base_url}auth#token={jwt_token}"
+
+        redirect_base_url = request.app.state.config.WEBUI_URL or request.base_url
+        if redirect_base_url.endswith("/"):
+            redirect_base_url = redirect_base_url[:-1]
+        redirect_url = f"{redirect_base_url}/auth#token={jwt_token}"
+
         return RedirectResponse(url=redirect_url, headers=response.headers)

backend/open_webui/utils/payload.py

@@ -1,5 +1,6 @@
 from open_webui.utils.task import prompt_template, prompt_variables_template
 from open_webui.utils.misc import (
+    deep_update,
     add_or_update_system_message,
 )
 
@@ -45,18 +46,29 @@ def apply_model_params_to_body(
     if not params:
         return form_data
 
-    for key, cast_func in mappings.items():
-        if (value := params.get(key)) is not None:
-            form_data[key] = cast_func(value)
+    for key, value in params.items():
+        if value is not None:
+            if key in mappings:
+                cast_func = mappings[key]
+                if isinstance(cast_func, Callable):
+                    form_data[key] = cast_func(value)
+            else:
+                form_data[key] = value
 
     return form_data
 
 
 # inplace function: form_data is modified
 def apply_model_params_to_body_openai(params: dict, form_data: dict) -> dict:
+    custom_params = params.pop("custom_params", {})
+    if custom_params:
+        # If there are custom parameters, we need to apply them first
+        params = deep_update(params, custom_params)
+
     mappings = {
         "temperature": float,
         "top_p": float,
+        "min_p": float,
         "max_tokens": int,
         "frequency_penalty": float,
         "presence_penalty": float,
@@ -70,6 +82,11 @@ def apply_model_params_to_body_openai(params: dict, form_data: dict) -> dict:
 
 
 def apply_model_params_to_body_ollama(params: dict, form_data: dict) -> dict:
+    custom_params = params.pop("custom_params", {})
+    if custom_params:
+        # If there are custom parameters, we need to apply them first
+        params = deep_update(params, custom_params)
+
     # Convert OpenAI parameter names to Ollama parameter names if needed.
     name_differences = {
         "max_tokens": "num_predict",

backend/open_webui/utils/plugin.py

@@ -115,7 +115,7 @@ def load_tool_module_by_id(tool_id, content=None):
         os.unlink(temp_file.name)
 
 
-def load_function_module_by_id(function_id, content=None):
+def load_function_module_by_id(function_id: str, content: str | None = None):
     if content is None:
         function = Functions.get_function_by_id(function_id)
         if not function:
@@ -166,6 +166,62 @@ def load_function_module_by_id(function_id, content=None):
         os.unlink(temp_file.name)
 
 
+def get_function_module_from_cache(request, function_id, load_from_db=True):
+    if load_from_db:
+        # Always load from the database by default
+        # This is useful for hooks like "inlet" or "outlet" where the content might change
+        # and we want to ensure the latest content is used.
+
+        function = Functions.get_function_by_id(function_id)
+        if not function:
+            raise Exception(f"Function not found: {function_id}")
+        content = function.content
+
+        new_content = replace_imports(content)
+        if new_content != content:
+            content = new_content
+            # Update the function content in the database
+            Functions.update_function_by_id(function_id, {"content": content})
+
+        if (
+            hasattr(request.app.state, "FUNCTION_CONTENTS")
+            and function_id in request.app.state.FUNCTION_CONTENTS
+        ) and (
+            hasattr(request.app.state, "FUNCTIONS")
+            and function_id in request.app.state.FUNCTIONS
+        ):
+            if request.app.state.FUNCTION_CONTENTS[function_id] == content:
+                return request.app.state.FUNCTIONS[function_id], None, None
+
+        function_module, function_type, frontmatter = load_function_module_by_id(
+            function_id, content
+        )
+    else:
+        # Load from cache (e.g. "stream" hook)
+        # This is useful for performance reasons
+
+        if (
+            hasattr(request.app.state, "FUNCTIONS")
+            and function_id in request.app.state.FUNCTIONS
+        ):
+            return request.app.state.FUNCTIONS[function_id], None, None
+
+        function_module, function_type, frontmatter = load_function_module_by_id(
+            function_id
+        )
+
+    if not hasattr(request.app.state, "FUNCTIONS"):
+        request.app.state.FUNCTIONS = {}
+
+    if not hasattr(request.app.state, "FUNCTION_CONTENTS"):
+        request.app.state.FUNCTION_CONTENTS = {}
+
+    request.app.state.FUNCTIONS[function_id] = function_module
+    request.app.state.FUNCTION_CONTENTS[function_id] = content
+
+    return function_module, function_type, frontmatter
+
+
 def install_frontmatter_requirements(requirements: str):
     if requirements:
         try:

backend/open_webui/utils/task.py

@@ -22,7 +22,7 @@ def get_task_model_id(
     # Set the task model
     task_model_id = default_model_id
     # Check if the user has a custom task model and use that model
-    if models[task_model_id].get("owned_by") == "ollama":
+    if models[task_model_id].get("connection_type") == "local":
         if task_model and task_model in models:
             task_model_id = task_model
     else:

backend/open_webui/utils/tools.py

@@ -37,6 +37,7 @@ from open_webui.models.tools import Tools
 from open_webui.models.users import UserModel
 from open_webui.utils.plugin import load_tool_module_by_id
 from open_webui.env import (
+    SRC_LOG_LEVELS,
     AIOHTTP_CLIENT_TIMEOUT_TOOL_SERVER_DATA,
     AIOHTTP_CLIENT_SESSION_TOOL_SERVER_SSL,
 )
@@ -44,6 +45,7 @@ from open_webui.env import (
 import copy
 
 log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["MODELS"])
 
 
 def get_async_tool_function_and_apply_extra_params(
@@ -158,7 +160,7 @@ def get_tools(
                 # TODO: Fix hack for OpenAI API
                 # Some times breaks OpenAI but others don't. Leaving the comment
                 for val in spec.get("parameters", {}).get("properties", {}).values():
-                    if val["type"] == "str":
+                    if val.get("type") == "str":
                         val["type"] = "string"
 
                 # Remove internal reserved parameters (e.g. __id__, __user__)
@@ -477,7 +479,7 @@ async def get_tool_server_data(token: str, url: str) -> Dict[str, Any]:
         "specs": convert_openapi_to_tool_payload(res),
     }
 
-    print("Fetched data:", data)
+    log.info("Fetched data:", data)
     return data
 
 
@@ -488,8 +490,19 @@ async def get_tool_servers_data(
     server_entries = []
     for idx, server in enumerate(servers):
         if server.get("config", {}).get("enable"):
-            url_path = server.get("path", "openapi.json")
-            full_url = f"{server.get('url')}/{url_path}"
+            # Path (to OpenAPI spec URL) can be either a full URL or a path to append to the base URL
+            openapi_path = server.get("path", "openapi.json")
+            if "://" in openapi_path:
+                # If it contains "://", it's a full URL
+                full_url = openapi_path
+            else:
+                if not openapi_path.startswith("/"):
+                    # Ensure the path starts with a slash
+                    openapi_path = f"/{openapi_path}"
+
+                full_url = f"{server.get('url')}{openapi_path}"
+
+            info = server.get("info", {})
 
             auth_type = server.get("auth_type", "bearer")
             token = None
@@ -498,26 +511,37 @@ async def get_tool_servers_data(
                 token = server.get("key", "")
             elif auth_type == "session":
                 token = session_token
-            server_entries.append((idx, server, full_url, token))
+            server_entries.append((idx, server, full_url, info, token))
 
     # Create async tasks to fetch data
-    tasks = [get_tool_server_data(token, url) for (_, _, url, token) in server_entries]
+    tasks = [
+        get_tool_server_data(token, url) for (_, _, url, _, token) in server_entries
+    ]
 
     # Execute tasks concurrently
     responses = await asyncio.gather(*tasks, return_exceptions=True)
 
     # Build final results with index and server metadata
     results = []
-    for (idx, server, url, _), response in zip(server_entries, responses):
+    for (idx, server, url, info, _), response in zip(server_entries, responses):
         if isinstance(response, Exception):
-            print(f"Failed to connect to {url} OpenAPI tool server")
+            log.error(f"Failed to connect to {url} OpenAPI tool server")
             continue
 
+        openapi_data = response.get("openapi", {})
+
+        if info and isinstance(openapi_data, dict):
+            if "name" in info:
+                openapi_data["info"]["title"] = info.get("name", "Tool Server")
+
+            if "description" in info:
+                openapi_data["info"]["description"] = info.get("description", "")
+
         results.append(
             {
                 "idx": idx,
                 "url": server.get("url"),
-                "openapi": response.get("openapi"),
+                "openapi": openapi_data,
                 "info": response.get("info"),
                 "specs": response.get("specs"),
             }
@@ -620,5 +644,5 @@ async def execute_tool_server(
 
     except Exception as err:
         error = str(err)
-        print("API Request Error:", error)
+        log.exception("API Request Error:", error)
         return {"error": error}

backend/requirements.txt

@@ -12,10 +12,12 @@ aiohttp==3.11.11
 async-timeout
 aiocache
 aiofiles
+starlette-compress==1.6.0
+
 
 sqlalchemy==2.0.38
 alembic==1.14.0
-peewee==3.17.9
+peewee==3.18.1
 peewee-migrate==1.12.2
 psycopg2-binary==2.9.9
 pgvector==0.4.0
@@ -37,7 +39,8 @@ asgiref==3.8.1
 # AI libraries
 openai
 anthropic
-google-generativeai==0.8.4
+google-genai==1.15.0
+google-generativeai==0.8.5
 tiktoken
 
 langchain==0.3.24
@@ -98,7 +101,7 @@ pytube==15.0.0
 
 extract_msg
 pydub
-duckduckgo-search~=8.0.0
+duckduckgo-search==8.0.2
 
 ## Google Drive
 google-api-python-client

package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "open-webui",
-	"version": "0.6.8",
+	"version": "0.6.12",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "open-webui",
-			"version": "0.6.8",
+			"version": "0.6.12",
 			"dependencies": {
 				"@azure/msal-browser": "^4.5.0",
 				"@codemirror/lang-javascript": "^6.2.2",
@@ -22,6 +22,10 @@
 				"@tiptap/extension-code-block-lowlight": "^2.11.9",
 				"@tiptap/extension-highlight": "^2.10.0",
 				"@tiptap/extension-placeholder": "^2.10.0",
+				"@tiptap/extension-table": "^2.12.0",
+				"@tiptap/extension-table-cell": "^2.12.0",
+				"@tiptap/extension-table-header": "^2.12.0",
+				"@tiptap/extension-table-row": "^2.12.0",
 				"@tiptap/extension-typography": "^2.10.0",
 				"@tiptap/pm": "^2.11.7",
 				"@tiptap/starter-kit": "^2.10.0",
@@ -62,6 +66,7 @@
 				"prosemirror-schema-basic": "^1.2.3",
 				"prosemirror-schema-list": "^1.5.1",
 				"prosemirror-state": "^1.4.3",
+				"prosemirror-tables": "^1.7.1",
 				"prosemirror-view": "^1.34.3",
 				"pyodide": "^0.27.3",
 				"socket.io-client": "^4.2.0",
@@ -69,6 +74,7 @@
 				"svelte-sonner": "^0.3.19",
 				"tippy.js": "^6.3.7",
 				"turndown": "^7.2.0",
+				"turndown-plugin-gfm": "^1.0.2",
 				"undici": "^7.3.0",
 				"uuid": "^9.0.1",
 				"vite-plugin-static-copy": "^2.2.0",
@@ -3173,6 +3179,59 @@
 				"@tiptap/core": "^2.7.0"
 			}
 		},
+		"node_modules/@tiptap/extension-table": {
+			"version": "2.12.0",
+			"resolved": "https://registry.npmjs.org/@tiptap/extension-table/-/extension-table-2.12.0.tgz",
+			"integrity": "sha512-tT3IbbBal0vPQ1Bc/3Xl+tmqqZQCYWxnycBPl/WZBqhd57DWzfJqRPESwCGUIJgjOtTnipy/ulvj0FxHi1j9JA==",
+			"license": "MIT",
+			"funding": {
+				"type": "github",
+				"url": "https://github.com/sponsors/ueberdosis"
+			},
+			"peerDependencies": {
+				"@tiptap/core": "^2.7.0",
+				"@tiptap/pm": "^2.7.0"
+			}
+		},
+		"node_modules/@tiptap/extension-table-cell": {
+			"version": "2.12.0",
+			"resolved": "https://registry.npmjs.org/@tiptap/extension-table-cell/-/extension-table-cell-2.12.0.tgz",
+			"integrity": "sha512-8i35uCkmkSiQxMiZ+DLgT/wj24P5U/Zo3jr1e0tMAAMG7sRO1MljjLmkpV8WCdBo0xoRqzkz4J7Nkq+DtzZv9Q==",
+			"license": "MIT",
+			"funding": {
+				"type": "github",
+				"url": "https://github.com/sponsors/ueberdosis"
+			},
+			"peerDependencies": {
+				"@tiptap/core": "^2.7.0"
+			}
+		},
+		"node_modules/@tiptap/extension-table-header": {
+			"version": "2.12.0",
+			"resolved": "https://registry.npmjs.org/@tiptap/extension-table-header/-/extension-table-header-2.12.0.tgz",
+			"integrity": "sha512-gRKEsy13KKLpg9RxyPeUGqh4BRFSJ2Bc2KQP1ldhef6CPRYHCbGycxXCVQ5aAb7Mhpo54L+AAkmAv1iMHUTflw==",
+			"license": "MIT",
+			"funding": {
+				"type": "github",
+				"url": "https://github.com/sponsors/ueberdosis"
+			},
+			"peerDependencies": {
+				"@tiptap/core": "^2.7.0"
+			}
+		},
+		"node_modules/@tiptap/extension-table-row": {
+			"version": "2.12.0",
+			"resolved": "https://registry.npmjs.org/@tiptap/extension-table-row/-/extension-table-row-2.12.0.tgz",
+			"integrity": "sha512-AEW/Zl9V0IoaYDBLMhF5lVl0xgoIJs3IuKCsIYxGDlxBfTVFC6PfQzvuy296CMjO5ZcZ0xalVipPV9ggsMRD+w==",
+			"license": "MIT",
+			"funding": {
+				"type": "github",
+				"url": "https://github.com/sponsors/ueberdosis"
+			},
+			"peerDependencies": {
+				"@tiptap/core": "^2.7.0"
+			}
+		},
 		"node_modules/@tiptap/extension-text": {
 			"version": "2.10.0",
 			"resolved": "https://registry.npmjs.org/@tiptap/extension-text/-/extension-text-2.10.0.tgz",
@@ -9809,16 +9868,16 @@
 			}
 		},
 		"node_modules/prosemirror-tables": {
-			"version": "1.6.4",
-			"resolved": "https://registry.npmjs.org/prosemirror-tables/-/prosemirror-tables-1.6.4.tgz",
-			"integrity": "sha512-TkDY3Gw52gRFRfRn2f4wJv5WOgAOXLJA2CQJYIJ5+kdFbfj3acR4JUW6LX2e1hiEBiUwvEhzH5a3cZ5YSztpIA==",
+			"version": "1.7.1",
+			"resolved": "https://registry.npmjs.org/prosemirror-tables/-/prosemirror-tables-1.7.1.tgz",
+			"integrity": "sha512-eRQ97Bf+i9Eby99QbyAiyov43iOKgWa7QCGly+lrDt7efZ1v8NWolhXiB43hSDGIXT1UXgbs4KJN3a06FGpr1Q==",
 			"license": "MIT",
 			"dependencies": {
 				"prosemirror-keymap": "^1.2.2",
-				"prosemirror-model": "^1.24.1",
+				"prosemirror-model": "^1.25.0",
 				"prosemirror-state": "^1.4.3",
-				"prosemirror-transform": "^1.10.2",
-				"prosemirror-view": "^1.37.2"
+				"prosemirror-transform": "^1.10.3",
+				"prosemirror-view": "^1.39.1"
 			}
 		},
 		"node_modules/prosemirror-trailing-node": {
@@ -9837,9 +9896,9 @@
 			}
 		},
 		"node_modules/prosemirror-transform": {
-			"version": "1.10.2",
-			"resolved": "https://registry.npmjs.org/prosemirror-transform/-/prosemirror-transform-1.10.2.tgz",
-			"integrity": "sha512-2iUq0wv2iRoJO/zj5mv8uDUriOHWzXRnOTVgCzSXnktS/2iQRa3UUQwVlkBlYZFtygw6Nh1+X4mGqoYBINn5KQ==",
+			"version": "1.10.4",
+			"resolved": "https://registry.npmjs.org/prosemirror-transform/-/prosemirror-transform-1.10.4.tgz",
+			"integrity": "sha512-pwDy22nAnGqNR1feOQKHxoFkkUtepoFAd3r2hbEDsnf4wp57kKA36hXsB3njA9FtONBEwSDnDeCiJe+ItD+ykw==",
 			"license": "MIT",
 			"dependencies": {
 				"prosemirror-model": "^1.21.0"
@@ -11808,6 +11867,12 @@
 				"@mixmark-io/domino": "^2.2.0"
 			}
 		},
+		"node_modules/turndown-plugin-gfm": {
+			"version": "1.0.2",
+			"resolved": "https://registry.npmjs.org/turndown-plugin-gfm/-/turndown-plugin-gfm-1.0.2.tgz",
+			"integrity": "sha512-vwz9tfvF7XN/jE0dGoBei3FXWuvll78ohzCZQuOb+ZjWrs3a0XhQVomJEb2Qh4VHTPNRO4GPZh0V7VRbiWwkRg==",
+			"license": "MIT"
+		},
 		"node_modules/tweetnacl": {
 			"version": "0.14.5",
 			"resolved": "https://registry.npmjs.org/tweetnacl/-/tweetnacl-0.14.5.tgz",

package.json

@@ -1,11 +1,12 @@
 {
 	"name": "open-webui",
-	"version": "0.6.8",
+	"version": "0.6.12",
 	"private": true,
 	"scripts": {
 		"dev": "npm run pyodide:fetch && vite dev --host",
 		"dev:5050": "npm run pyodide:fetch && vite dev --port 5050",
 		"build": "npm run pyodide:fetch && vite build",
+		"build:watch": "npm run pyodide:fetch && vite build --watch",
 		"preview": "vite preview",
 		"check": "svelte-kit sync && svelte-check --tsconfig ./tsconfig.json",
 		"check:watch": "svelte-kit sync && svelte-check --tsconfig ./tsconfig.json --watch",
@@ -65,6 +66,10 @@
 		"@tiptap/extension-code-block-lowlight": "^2.11.9",
 		"@tiptap/extension-highlight": "^2.10.0",
 		"@tiptap/extension-placeholder": "^2.10.0",
+		"@tiptap/extension-table": "^2.12.0",
+		"@tiptap/extension-table-cell": "^2.12.0",
+		"@tiptap/extension-table-header": "^2.12.0",
+		"@tiptap/extension-table-row": "^2.12.0",
 		"@tiptap/extension-typography": "^2.10.0",
 		"@tiptap/pm": "^2.11.7",
 		"@tiptap/starter-kit": "^2.10.0",
@@ -105,6 +110,7 @@
 		"prosemirror-schema-basic": "^1.2.3",
 		"prosemirror-schema-list": "^1.5.1",
 		"prosemirror-state": "^1.4.3",
+		"prosemirror-tables": "^1.7.1",
 		"prosemirror-view": "^1.34.3",
 		"pyodide": "^0.27.3",
 		"socket.io-client": "^4.2.0",
@@ -112,6 +118,7 @@
 		"svelte-sonner": "^0.3.19",
 		"tippy.js": "^6.3.7",
 		"turndown": "^7.2.0",
+		"turndown-plugin-gfm": "^1.0.2",
 		"undici": "^7.3.0",
 		"uuid": "^9.0.1",
 		"vite-plugin-static-copy": "^2.2.0",

pyproject.toml

@@ -21,9 +21,11 @@ dependencies = [
     "aiocache",
     "aiofiles",
 
+    "starlette-compress==1.6.0",
+
     "sqlalchemy==2.0.38",
     "alembic==1.14.0",
-    "peewee==3.17.9",
+    "peewee==3.18.1",
     "peewee-migrate==1.12.2",
     "psycopg2-binary==2.9.9",
     "pgvector==0.4.0",
@@ -45,7 +47,8 @@ dependencies = [
 
     "openai",
     "anthropic",
-    "google-generativeai==0.8.4",
+    "google-genai==1.15.0",
+    "google-generativeai==0.8.5",
     "tiktoken",
 
     "langchain==0.3.24",
@@ -105,7 +108,7 @@ dependencies = [
 
     "extract_msg",
     "pydub",
-    "duckduckgo-search~=8.0.0",
+    "duckduckgo-search==8.0.2",
 
     "google-api-python-client",
     "google-auth-httplib2",

src/app.css

@@ -81,7 +81,7 @@ textarea::placeholder {
 }
 
 .font-primary {
-	font-family: 'Archivo', sans-serif;
+	font-family: 'Archivo', 'Vazirmatn', sans-serif;
 }
 
 .drag-region {
@@ -103,7 +103,7 @@ li p {
 
 ::-webkit-scrollbar-thumb {
 	--tw-border-opacity: 1;
-	background-color: rgba(236, 236, 236, 0.8);
+	background-color: rgba(215, 215, 215, 0.8);
 	border-color: rgba(255, 255, 255, var(--tw-border-opacity));
 	border-radius: 9999px;
 	border-width: 1px;
@@ -111,12 +111,12 @@ li p {
 
 /* Dark theme scrollbar styles */
 .dark ::-webkit-scrollbar-thumb {
-	background-color: rgba(42, 42, 42, 0.8); /* Darker color for dark theme */
+	background-color: rgba(67, 67, 67, 0.8); /* Darker color for dark theme */
 	border-color: rgba(0, 0, 0, var(--tw-border-opacity));
 }
 
 ::-webkit-scrollbar {
-	height: 0.8rem;
+	height: 0.6rem;
 	width: 0.4rem;
 }
 
@@ -314,12 +314,20 @@ input[type='number'] {
 .ProseMirror p.is-editor-empty:first-child::before {
 	content: attr(data-placeholder);
 	float: left;
-	color: #adb5bd;
+	/* Below color is from tailwind, and has the proper contrast
+	text-gray-600 from: https://tailwindcss.com/docs/color */
+	color: #676767;
 	pointer-events: none;
 
 	@apply line-clamp-1 absolute;
 }
 
+@media (prefers-color-scheme: dark) {
+	.ProseMirror p.is-editor-empty:first-child::before {
+		color: #757575;
+	}
+}
+
 .ai-autocompletion::after {
 	color: #a0a0a0;
 
@@ -404,3 +412,29 @@ input[type='number'] {
 .hljs-strong {
 	font-weight: 700;
 }
+
+/* Table styling for tiptap editors */
+.tiptap table {
+	@apply w-full text-sm text-left text-gray-500 dark:text-gray-400 max-w-full;
+}
+
+.tiptap thead {
+	@apply text-xs text-gray-700 uppercase bg-gray-50 dark:bg-gray-850 dark:text-gray-400 border-none;
+}
+
+.tiptap th,
+.tiptap td {
+	@apply px-3 py-1.5 border border-gray-100 dark:border-gray-850;
+}
+
+.tiptap th {
+	@apply cursor-pointer text-left text-xs text-gray-700 dark:text-gray-400 font-semibold uppercase bg-gray-50 dark:bg-gray-850;
+}
+
+.tiptap td {
+	@apply text-gray-900 dark:text-white w-max;
+}
+
+.tiptap tr {
+	@apply bg-white dark:bg-gray-900 dark:border-gray-850 text-xs;
+}

src/lib/apis/audio/index.ts

@@ -15,7 +15,7 @@ export const getAudioConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -52,7 +52,7 @@ export const updateAudioConfig = async (token: string, payload: OpenAIConfigForm
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -64,9 +64,12 @@ export const updateAudioConfig = async (token: string, payload: OpenAIConfigForm
 	return res;
 };
 
-export const transcribeAudio = async (token: string, file: File) => {
+export const transcribeAudio = async (token: string, file: File, language?: string) => {
 	const data = new FormData();
 	data.append('file', file);
+	if (language) {
+		data.append('language', language);
+	}
 
 	let error = null;
 	const res = await fetch(`${AUDIO_API_BASE_URL}/transcriptions`, {
@@ -83,7 +86,7 @@ export const transcribeAudio = async (token: string, file: File) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -120,7 +123,7 @@ export const synthesizeOpenAISpeech = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 
 			return null;
 		});
@@ -152,7 +155,7 @@ export const getModels = async (token: string = ''): Promise<AvailableModelsResp
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 
 			return null;
 		});
@@ -180,7 +183,7 @@ export const getVoices = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 
 			return null;
 		});

src/lib/apis/auths/index.ts

@@ -15,7 +15,7 @@ export const getAdminDetails = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -42,7 +42,7 @@ export const getAdminConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -70,7 +70,7 @@ export const updateAdminConfig = async (token: string, body: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -98,7 +98,7 @@ export const getSessionUser = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -129,7 +129,7 @@ export const ldapUserSignIn = async (user: string, password: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 
 			error = err.detail;
 			return null;
@@ -157,7 +157,7 @@ export const getLdapConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -187,7 +187,7 @@ export const updateLdapConfig = async (token: string = '', enable_ldap: boolean)
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -214,7 +214,7 @@ export const getLdapServer = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -242,7 +242,7 @@ export const updateLdapServer = async (token: string = '', body: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -273,7 +273,7 @@ export const userSignIn = async (email: string, password: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 
 			error = err.detail;
 			return null;
@@ -312,7 +312,7 @@ export const userSignUp = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -339,7 +339,7 @@ export const userSignOut = async () => {
 			return res;
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -347,6 +347,7 @@ export const userSignOut = async () => {
 	if (error) {
 		throw error;
 	}
+	return res;
 };
 
 export const addUser = async (
@@ -378,7 +379,7 @@ export const addUser = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -409,7 +410,7 @@ export const updateUserProfile = async (token: string, name: string, profileImag
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -440,7 +441,7 @@ export const updateUserPassword = async (token: string, password: string, newPas
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -467,7 +468,7 @@ export const getSignUpEnabledStatus = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -494,7 +495,7 @@ export const getDefaultUserRole = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -524,7 +525,7 @@ export const updateDefaultUserRole = async (token: string, role: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -551,7 +552,7 @@ export const toggleSignUpEnabledStatus = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -578,7 +579,7 @@ export const getJWTExpiresDuration = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -608,7 +609,7 @@ export const updateJWTExpiresDuration = async (token: string, duration: string)
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -635,7 +636,7 @@ export const createAPIKey = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -660,7 +661,7 @@ export const getAPIKey = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -685,7 +686,7 @@ export const deleteAPIKey = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});

src/lib/apis/channels/index.ts

@@ -28,7 +28,7 @@ export const createNewChannel = async (token: string = '', channel: ChannelForm)
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -59,7 +59,7 @@ export const getChannels = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -90,7 +90,7 @@ export const getChannelById = async (token: string = '', channel_id: string) =>
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -126,7 +126,7 @@ export const updateChannelById = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -157,7 +157,7 @@ export const deleteChannelById = async (token: string = '', channel_id: string)
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -196,7 +196,7 @@ export const getChannelMessages = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -236,7 +236,7 @@ export const getChannelThreadMessages = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -275,7 +275,7 @@ export const sendMessage = async (token: string = '', channel_id: string, messag
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -315,7 +315,7 @@ export const updateMessage = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -355,7 +355,7 @@ export const addReaction = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -395,7 +395,7 @@ export const removeReaction = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -429,7 +429,7 @@ export const deleteMessage = async (token: string = '', channel_id: string, mess
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/chats/index.ts

@@ -21,7 +21,7 @@ export const createNewChat = async (token: string, chat: object) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -61,7 +61,7 @@ export const importChat = async (
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -97,7 +97,7 @@ export const getChatList = async (token: string = '', page: number | null = null
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -111,10 +111,79 @@ export const getChatList = async (token: string = '', page: number | null = null
 	}));
 };
 
-export const getChatListByUserId = async (token: string = '', userId: string) => {
+export const getChatListByUserId = async (
+	token: string = '',
+	userId: string,
+	page: number = 1,
+	filter?: object
+) => {
 	let error = null;
 
-	const res = await fetch(`${WEBUI_API_BASE_URL}/chats/list/user/${userId}`, {
+	const searchParams = new URLSearchParams();
+
+	searchParams.append('page', `${page}`);
+
+	if (filter) {
+		Object.entries(filter).forEach(([key, value]) => {
+			if (value !== undefined && value !== null) {
+				searchParams.append(key, value.toString());
+			}
+		});
+	}
+
+	const res = await fetch(
+		`${WEBUI_API_BASE_URL}/chats/list/user/${userId}?${searchParams.toString()}`,
+		{
+			method: 'GET',
+			headers: {
+				Accept: 'application/json',
+				'Content-Type': 'application/json',
+				...(token && { authorization: `Bearer ${token}` })
+			}
+		}
+	)
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err;
+			console.error(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res.map((chat) => ({
+		...chat,
+		time_range: getTimeRange(chat.updated_at)
+	}));
+};
+
+export const getArchivedChatList = async (
+	token: string = '',
+	page: number = 1,
+	filter?: object
+) => {
+	let error = null;
+
+	const searchParams = new URLSearchParams();
+	searchParams.append('page', `${page}`);
+
+	if (filter) {
+		Object.entries(filter).forEach(([key, value]) => {
+			if (value !== undefined && value !== null) {
+				searchParams.append(key, value.toString());
+			}
+		});
+	}
+
+	const res = await fetch(`${WEBUI_API_BASE_URL}/chats/archived?${searchParams.toString()}`, {
 		method: 'GET',
 		headers: {
 			Accept: 'application/json',
@@ -131,7 +200,7 @@ export const getChatListByUserId = async (token: string = '', userId: string) =>
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -145,37 +214,6 @@ export const getChatListByUserId = async (token: string = '', userId: string) =>
 	}));
 };
 
-export const getArchivedChatList = async (token: string = '') => {
-	let error = null;
-
-	const res = await fetch(`${WEBUI_API_BASE_URL}/chats/archived`, {
-		method: 'GET',
-		headers: {
-			Accept: 'application/json',
-			'Content-Type': 'application/json',
-			...(token && { authorization: `Bearer ${token}` })
-		}
-	})
-		.then(async (res) => {
-			if (!res.ok) throw await res.json();
-			return res.json();
-		})
-		.then((json) => {
-			return json;
-		})
-		.catch((err) => {
-			error = err;
-			console.log(err);
-			return null;
-		});
-
-	if (error) {
-		throw error;
-	}
-
-	return res;
-};
-
 export const getAllChats = async (token: string) => {
 	let error = null;
 
@@ -196,7 +234,7 @@ export const getAllChats = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -231,7 +269,7 @@ export const getChatListBySearchText = async (token: string, text: string, page:
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -265,7 +303,7 @@ export const getChatsByFolderId = async (token: string, folderId: string) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -296,7 +334,7 @@ export const getAllArchivedChats = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -327,7 +365,7 @@ export const getAllUserChats = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -358,7 +396,7 @@ export const getAllTags = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -389,7 +427,7 @@ export const getPinnedChatList = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -426,7 +464,7 @@ export const getChatListByTagName = async (token: string = '', tagName: string)
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -461,7 +499,7 @@ export const getChatById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -493,7 +531,7 @@ export const getChatByShareId = async (token: string, share_id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -531,7 +569,7 @@ export const getChatPinnedStatusById = async (token: string, id: string) => {
 				error = err;
 			}
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -569,7 +607,7 @@ export const toggleChatPinnedStatusById = async (token: string, id: string) => {
 				error = err;
 			}
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -610,7 +648,7 @@ export const cloneChatById = async (token: string, id: string, title?: string) =
 				error = err;
 			}
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -648,7 +686,7 @@ export const cloneSharedChatById = async (token: string, id: string) => {
 				error = err;
 			}
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -680,7 +718,7 @@ export const shareChatById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -715,7 +753,7 @@ export const updateChatFolderIdById = async (token: string, id: string, folderId
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -747,7 +785,7 @@ export const archiveChatById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -779,7 +817,7 @@ export const deleteSharedChatById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -814,7 +852,7 @@ export const updateChatById = async (token: string, id: string, chat: object) =>
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -846,7 +884,7 @@ export const deleteChatById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -878,7 +916,7 @@ export const getTagsById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -912,7 +950,7 @@ export const addTagById = async (token: string, id: string, tagName: string) =>
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -947,7 +985,7 @@ export const deleteTagById = async (token: string, id: string, tagName: string)
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -978,7 +1016,7 @@ export const deleteTagsById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -1010,7 +1048,7 @@ export const deleteAllChats = async (token: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -1042,7 +1080,7 @@ export const archiveAllChats = async (token: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/configs/index.ts

@@ -19,7 +19,7 @@ export const importConfig = async (token: string, config) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -46,7 +46,7 @@ export const exportConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -73,7 +73,7 @@ export const getDirectConnectionsConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -103,7 +103,7 @@ export const setDirectConnectionsConfig = async (token: string, config: object)
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -130,7 +130,7 @@ export const getToolServerConnections = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -160,7 +160,7 @@ export const setToolServerConnections = async (token: string, connections: objec
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -190,7 +190,7 @@ export const verifyToolServerConnection = async (token: string, connection: obje
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -217,7 +217,7 @@ export const getCodeExecutionConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -247,7 +247,7 @@ export const setCodeExecutionConfig = async (token: string, config: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -274,7 +274,7 @@ export const getModelsConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -304,7 +304,7 @@ export const setModelsConfig = async (token: string, config: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -334,7 +334,7 @@ export const setDefaultPromptSuggestions = async (token: string, promptSuggestio
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -361,7 +361,7 @@ export const getBanners = async (token: string): Promise<Banner[]> => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -391,7 +391,7 @@ export const setBanners = async (token: string, banners: Banner[]) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});

src/lib/apis/evaluations/index.ts

@@ -20,7 +20,7 @@ export const getConfig = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -51,7 +51,7 @@ export const updateConfig = async (token: string, config: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -82,7 +82,7 @@ export const getAllFeedbacks = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -113,7 +113,7 @@ export const exportAllFeedbacks = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -144,7 +144,7 @@ export const createNewFeedback = async (token: string, feedback: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -175,7 +175,7 @@ export const getFeedbackById = async (token: string, feedbackId: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -206,7 +206,7 @@ export const updateFeedbackById = async (token: string, feedbackId: string, feed
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -234,7 +234,7 @@ export const deleteFeedbackById = async (token: string, feedbackId: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/files/index.ts

@@ -1,8 +1,12 @@
 import { WEBUI_API_BASE_URL } from '$lib/constants';
 
-export const uploadFile = async (token: string, file: File) => {
+export const uploadFile = async (token: string, file: File, metadata?: object | null) => {
 	const data = new FormData();
 	data.append('file', file);
+	if (metadata) {
+		data.append('metadata', JSON.stringify(metadata));
+	}
+
 	let error = null;
 
 	const res = await fetch(`${WEBUI_API_BASE_URL}/files/`, {
@@ -19,7 +23,7 @@ export const uploadFile = async (token: string, file: File) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -76,7 +80,7 @@ export const getFiles = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -107,7 +111,7 @@ export const getFileById = async (token: string, id: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -141,7 +145,7 @@ export const updateFileDataContentById = async (token: string, id: string, conte
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -168,7 +172,7 @@ export const getFileContentById = async (id: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 
 			return null;
 		});
@@ -200,7 +204,7 @@ export const deleteFileById = async (token: string, id: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -231,7 +235,7 @@ export const deleteAllFiles = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/folders/index.ts

@@ -50,7 +50,7 @@ export const getFolders = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -81,7 +81,7 @@ export const getFolderById = async (token: string, id: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -115,7 +115,7 @@ export const updateFolderNameById = async (token: string, id: string, name: stri
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -153,7 +153,7 @@ export const updateFolderIsExpandedById = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -187,7 +187,7 @@ export const updateFolderParentIdById = async (token: string, id: string, parent
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -226,7 +226,7 @@ export const updateFolderItemsById = async (token: string, id: string, items: Fo
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -257,7 +257,7 @@ export const deleteFolderById = async (token: string, id: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/functions/index.ts

@@ -20,7 +20,7 @@ export const createNewFunction = async (token: string, func: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -51,7 +51,41 @@ export const getFunctions = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
+export const loadFunctionByUrl = async (token: string = '', url: string) => {
+	let error = null;
+
+	const res = await fetch(`${WEBUI_API_BASE_URL}/functions/load/url`, {
+		method: 'POST',
+		headers: {
+			Accept: 'application/json',
+			'Content-Type': 'application/json',
+			authorization: `Bearer ${token}`
+		},
+		body: JSON.stringify({
+			url
+		})
+	})
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err.detail;
+			console.error(err);
 			return null;
 		});
 
@@ -82,7 +116,7 @@ export const exportFunctions = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -114,7 +148,7 @@ export const getFunctionById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -149,7 +183,7 @@ export const updateFunctionById = async (token: string, id: string, func: object
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -181,7 +215,7 @@ export const deleteFunctionById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -213,7 +247,7 @@ export const toggleFunctionById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -245,7 +279,7 @@ export const toggleGlobalById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -277,7 +311,7 @@ export const getFunctionValvesById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -309,7 +343,7 @@ export const getFunctionValvesSpecById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -344,7 +378,7 @@ export const updateFunctionValvesById = async (token: string, id: string, valves
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -376,7 +410,7 @@ export const getUserValvesById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -408,7 +442,7 @@ export const getUserValvesSpecById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -443,7 +477,7 @@ export const updateUserValvesById = async (token: string, id: string, valves: ob
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/groups/index.ts

@@ -20,7 +20,7 @@ export const createNewGroup = async (token: string, group: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -51,7 +51,7 @@ export const getGroups = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -83,7 +83,7 @@ export const getGroupById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -118,7 +118,7 @@ export const updateGroupById = async (token: string, id: string, group: object)
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -150,7 +150,7 @@ export const deleteGroupById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/images/index.ts

@@ -16,7 +16,7 @@ export const getConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -51,7 +51,7 @@ export const updateConfig = async (token: string = '', config: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -83,7 +83,7 @@ export const verifyConfigUrl = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -115,7 +115,7 @@ export const getImageGenerationConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -148,7 +148,7 @@ export const updateImageGenerationConfig = async (token: string = '', config: ob
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -180,7 +180,7 @@ export const getImageGenerationModels = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -215,7 +215,7 @@ export const imageGenerations = async (token: string = '', prompt: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {

src/lib/apis/index.ts

@@ -25,7 +25,7 @@ export const getModels = async (
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -173,7 +173,7 @@ export const chatCompleted = async (token: string, body: ChatCompletedForm) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -212,7 +212,7 @@ export const chatAction = async (token: string, action_id: string, body: ChatAct
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -244,7 +244,7 @@ export const stopTask = async (token: string, id: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -276,7 +276,7 @@ export const getTaskIdsByChatId = async (token: string, chat_id: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -315,7 +315,7 @@ export const getToolServerData = async (token: string, url: string) => {
 			}
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -346,11 +346,15 @@ export const getToolServersData = async (i18n, servers: object[]) => {
 				.map(async (server) => {
 					const data = await getToolServerData(
 						(server?.auth_type ?? 'bearer') === 'bearer' ? server?.key : localStorage.token,
-						server?.url + '/' + (server?.path ?? 'openapi.json')
+						(server?.path ?? '').includes('://')
+							? server?.path
+							: `${server?.url}${(server?.path ?? '').startsWith('/') ? '' : '/'}${server?.path}`
 					).catch((err) => {
 						toast.error(
 							i18n.t(`Failed to connect to {{URL}} OpenAPI tool server`, {
-								URL: server?.url + '/' + (server?.path ?? 'openapi.json')
+								URL: (server?.path ?? '').includes('://')
+									? server?.path
+									: `${server?.url}${(server?.path ?? '').startsWith('/') ? '' : '/'}${server?.path}`
 							})
 						);
 						return null;
@@ -491,7 +495,7 @@ export const getTaskConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -520,7 +524,7 @@ export const updateTaskConfig = async (token: string, config: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -562,7 +566,7 @@ export const generateTitle = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			}
@@ -634,7 +638,7 @@ export const generateTags = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			}
@@ -706,7 +710,7 @@ export const generateEmoji = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			}
@@ -756,7 +760,7 @@ export const generateQueries = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			}
@@ -828,7 +832,7 @@ export const generateAutoCompletion = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			}
@@ -892,7 +896,7 @@ export const generateMoACompletion = async (
 			stream: true
 		})
 	}).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 		return null;
 	});
@@ -920,7 +924,7 @@ export const getPipelinesList = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -954,7 +958,7 @@ export const uploadPipeline = async (token: string, file: File, urlIdx: string)
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -990,7 +994,7 @@ export const downloadPipeline = async (token: string, url: string, urlIdx: strin
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -1026,7 +1030,7 @@ export const deletePipeline = async (token: string, id: string, urlIdx: string)
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -1063,7 +1067,7 @@ export const getPipelines = async (token: string, urlIdx?: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1100,7 +1104,7 @@ export const getPipelineValves = async (token: string, pipeline_id: string, urlI
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1136,7 +1140,7 @@ export const getPipelineValvesSpec = async (token: string, pipeline_id: string,
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1178,7 +1182,7 @@ export const updatePipelineValves = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 
 			if ('detail' in err) {
 				error = err.detail;
@@ -1210,7 +1214,7 @@ export const getBackendConfig = async () => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1236,7 +1240,7 @@ export const getChangelog = async () => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1263,7 +1267,7 @@ export const getVersionUpdates = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1290,7 +1294,7 @@ export const getModelFilterConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1325,7 +1329,7 @@ export const updateModelFilterConfig = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1352,7 +1356,7 @@ export const getWebhookUrl = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1382,7 +1386,7 @@ export const updateWebhookUrl = async (token: string, url: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1409,7 +1413,7 @@ export const getCommunitySharingEnabledStatus = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1436,7 +1440,7 @@ export const toggleCommunitySharingEnabledStatus = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -1463,7 +1467,7 @@ export const getModelConfig = async (token: string): Promise<GlobalModelConfig>
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -1511,7 +1515,7 @@ export const updateModelConfig = async (token: string, config: GlobalModelConfig
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});

src/lib/apis/knowledge/index.ts

@@ -27,7 +27,7 @@ export const createNewKnowledge = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -58,7 +58,7 @@ export const getKnowledgeBases = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -89,7 +89,7 @@ export const getKnowledgeBaseList = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -121,7 +121,7 @@ export const getKnowledgeById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -166,7 +166,7 @@ export const updateKnowledgeById = async (token: string, id: string, form: Knowl
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -201,7 +201,7 @@ export const addFileToKnowledgeById = async (token: string, id: string, fileId:
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -236,7 +236,7 @@ export const updateFileFromKnowledgeById = async (token: string, id: string, fil
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -271,7 +271,7 @@ export const removeFileFromKnowledgeById = async (token: string, id: string, fil
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -303,7 +303,7 @@ export const resetKnowledgeById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -335,7 +335,7 @@ export const deleteKnowledgeById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -363,7 +363,7 @@ export const reindexKnowledgeFiles = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/memories/index.ts

@@ -17,7 +17,7 @@ export const getMemories = async (token: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -48,7 +48,7 @@ export const addNewMemory = async (token: string, content: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -79,7 +79,7 @@ export const updateMemoryById = async (token: string, id: string, content: strin
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -110,7 +110,7 @@ export const queryMemory = async (token: string, content: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -142,7 +142,7 @@ export const deleteMemoryById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -174,7 +174,7 @@ export const deleteMemoriesByUserId = async (token: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/models/index.ts

@@ -20,7 +20,7 @@ export const getModels = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -51,7 +51,7 @@ export const getBaseModels = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -80,7 +80,7 @@ export const createNewModel = async (token: string, model: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -115,7 +115,7 @@ export const getModelById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -150,7 +150,7 @@ export const toggleModelById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -186,7 +186,7 @@ export const updateModelById = async (token: string, id: string, model: object)
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -221,7 +221,7 @@ export const deleteModelById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -253,7 +253,7 @@ export const deleteAllModels = async (token: string) => {
 		.catch((err) => {
 			error = err;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/notes/index.ts

@@ -28,7 +28,7 @@ export const createNewNote = async (token: string, note: NoteItem) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -59,7 +59,7 @@ export const getNotes = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -108,7 +108,7 @@ export const getNoteById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -143,7 +143,7 @@ export const updateNoteById = async (token: string, id: string, note: NoteItem)
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -175,7 +175,7 @@ export const deleteNoteById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/ollama/index.ts

@@ -1,10 +1,6 @@
 import { OLLAMA_API_BASE_URL } from '$lib/constants';
 
-export const verifyOllamaConnection = async (
-	token: string = '',
-	url: string = '',
-	key: string = ''
-) => {
+export const verifyOllamaConnection = async (token: string = '', connection: dict = {}) => {
 	let error = null;
 
 	const res = await fetch(`${OLLAMA_API_BASE_URL}/verify`, {
@@ -15,8 +11,7 @@ export const verifyOllamaConnection = async (
 			'Content-Type': 'application/json'
 		},
 		body: JSON.stringify({
-			url,
-			key
+			...connection
 		})
 	})
 		.then(async (res) => {
@@ -51,7 +46,7 @@ export const getOllamaConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -92,7 +87,7 @@ export const updateOllamaConfig = async (token: string = '', config: OllamaConfi
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -124,7 +119,7 @@ export const getOllamaUrls = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -159,7 +154,7 @@ export const updateOllamaUrls = async (token: string = '', urls: string[]) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -191,7 +186,7 @@ export const getOllamaVersion = async (token: string, urlIdx?: number) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -223,7 +218,7 @@ export const getOllamaModels = async (token: string = '', urlIdx: null | number
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -268,7 +263,7 @@ export const generatePrompt = async (token: string = '', model: string, conversa
 			`
 		})
 	}).catch((err) => {
-		console.log(err);
+		console.error(err);
 		if ('detail' in err) {
 			error = err.detail;
 		}
@@ -360,6 +355,31 @@ export const generateChatCompletion = async (token: string = '', body: object) =
 	return [res, controller];
 };
 
+export const unloadModel = async (token: string, tagName: string) => {
+	let error = null;
+
+	const res = await fetch(`${OLLAMA_API_BASE_URL}/api/unload`, {
+		method: 'POST',
+		headers: {
+			Accept: 'application/json',
+			'Content-Type': 'application/json',
+			Authorization: `Bearer ${token}`
+		},
+		body: JSON.stringify({
+			name: tagName
+		})
+	}).catch((err) => {
+		error = err;
+		return null;
+	});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
 export const createModel = async (token: string, payload: object, urlIdx: string | null = null) => {
 	let error = null;
 
@@ -408,11 +428,11 @@ export const deleteModel = async (token: string, tagName: string, urlIdx: string
 			return res.json();
 		})
 		.then((json) => {
-			console.log(json);
+			console.debug(json);
 			return true;
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 
 			if ('detail' in err) {
@@ -445,7 +465,7 @@ export const pullModel = async (token: string, tagName: string, urlIdx: number |
 			name: tagName
 		})
 	}).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 
 		if ('detail' in err) {
@@ -481,7 +501,7 @@ export const downloadModel = async (
 			})
 		}
 	).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 
 		if ('detail' in err) {
@@ -512,7 +532,7 @@ export const uploadModel = async (token: string, file: File, urlIdx: string | nu
 			body: formData
 		}
 	).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 
 		if ('detail' in err) {

src/lib/apis/openai/index.ts

@@ -16,7 +16,7 @@ export const getOpenAIConfig = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -58,7 +58,7 @@ export const updateOpenAIConfig = async (token: string = '', config: OpenAIConfi
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -90,7 +90,7 @@ export const getOpenAIUrls = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -125,7 +125,7 @@ export const updateOpenAIUrls = async (token: string = '', urls: string[]) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -157,7 +157,7 @@ export const getOpenAIKeys = async (token: string = '') => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -192,7 +192,7 @@ export const updateOpenAIKeys = async (token: string = '', keys: string[]) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			if ('detail' in err) {
 				error = err.detail;
 			} else {
@@ -267,10 +267,10 @@ export const getOpenAIModels = async (token: string, urlIdx?: number) => {
 
 export const verifyOpenAIConnection = async (
 	token: string = '',
-	url: string = 'https://api.openai.com/v1',
-	key: string = '',
+	connection: dict = {},
 	direct: boolean = false
 ) => {
+	const { url, key, config } = connection;
 	if (!url) {
 		throw 'OpenAI: URL is required';
 	}
@@ -309,7 +309,8 @@ export const verifyOpenAIConnection = async (
 			},
 			body: JSON.stringify({
 				url,
-				key
+				key,
+				config
 			})
 		})
 			.then(async (res) => {
@@ -346,7 +347,7 @@ export const chatCompletion = async (
 		},
 		body: JSON.stringify(body)
 	}).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 		return null;
 	});
@@ -409,7 +410,7 @@ export const synthesizeOpenAISpeech = async (
 			voice: speaker
 		})
 	}).catch((err) => {
-		console.log(err);
+		console.error(err);
 		error = err;
 		return null;
 	});

src/lib/apis/prompts/index.ts

@@ -28,7 +28,7 @@ export const createNewPrompt = async (token: string, prompt: PromptItem) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -59,7 +59,7 @@ export const getPrompts = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -90,7 +90,7 @@ export const getPromptList = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -122,7 +122,7 @@ export const getPromptByCommand = async (token: string, command: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -158,7 +158,7 @@ export const updatePromptByCommand = async (token: string, prompt: PromptItem) =
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -192,7 +192,7 @@ export const deletePromptByCommand = async (token: string, command: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/retrieval/index.ts

@@ -15,7 +15,7 @@ export const getRAGConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -77,7 +77,7 @@ export const updateRAGConfig = async (token: string, payload: RAGConfigForm) =>
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -104,7 +104,7 @@ export const getQuerySettings = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -140,7 +140,7 @@ export const updateQuerySettings = async (token: string, settings: QuerySettings
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -167,7 +167,7 @@ export const getEmbeddingConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -209,7 +209,7 @@ export const updateEmbeddingConfig = async (token: string, payload: EmbeddingMod
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -236,7 +236,7 @@ export const getRerankingConfig = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -270,7 +270,7 @@ export const updateRerankingConfig = async (token: string, payload: RerankingMod
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -313,7 +313,7 @@ export const processFile = async (
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -344,7 +344,7 @@ export const processYoutubeVideo = async (token: string, url: string) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -376,7 +376,7 @@ export const processWeb = async (token: string, collection_name: string, url: st
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -410,7 +410,7 @@ export const processWebSearch = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});

src/lib/apis/tools/index.ts

@@ -20,7 +20,41 @@ export const createNewTool = async (token: string, tool: object) => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
+export const loadToolByUrl = async (token: string = '', url: string) => {
+	let error = null;
+
+	const res = await fetch(`${WEBUI_API_BASE_URL}/tools/load/url`, {
+		method: 'POST',
+		headers: {
+			Accept: 'application/json',
+			'Content-Type': 'application/json',
+			authorization: `Bearer ${token}`
+		},
+		body: JSON.stringify({
+			url
+		})
+	})
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err.detail;
+			console.error(err);
 			return null;
 		});
 
@@ -51,7 +85,7 @@ export const getTools = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -82,7 +116,7 @@ export const getToolList = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -113,7 +147,7 @@ export const exportTools = async (token: string = '') => {
 		})
 		.catch((err) => {
 			error = err.detail;
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -145,7 +179,7 @@ export const getToolById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -180,7 +214,7 @@ export const updateToolById = async (token: string, id: string, tool: object) =>
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -212,7 +246,7 @@ export const deleteToolById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -244,7 +278,7 @@ export const getToolValvesById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -276,7 +310,7 @@ export const getToolValvesSpecById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -311,7 +345,7 @@ export const updateToolValvesById = async (token: string, id: string, valves: ob
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -343,7 +377,7 @@ export const getUserValvesById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -375,7 +409,7 @@ export const getUserValvesSpecById = async (token: string, id: string) => {
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 
@@ -410,7 +444,7 @@ export const updateUserValvesById = async (token: string, id: string, valves: ob
 		.catch((err) => {
 			error = err.detail;
 
-			console.log(err);
+			console.error(err);
 			return null;
 		});
 

src/lib/apis/users/index.ts

@@ -16,7 +16,7 @@ export const getUserGroups = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -43,7 +43,7 @@ export const getUserDefaultPermissions = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -73,7 +73,7 @@ export const updateUserDefaultPermissions = async (token: string, permissions: o
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -104,7 +104,7 @@ export const updateUserRole = async (token: string, id: string, role: string) =>
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -154,7 +154,7 @@ export const getUsers = async (
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -182,7 +182,7 @@ export const getAllUsers = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -208,7 +208,7 @@ export const getUserSettings = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -238,7 +238,7 @@ export const updateUserSettings = async (token: string, settings: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -265,7 +265,7 @@ export const getUserById = async (token: string, userId: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -291,7 +291,7 @@ export const getUserInfo = async (token: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -321,7 +321,7 @@ export const updateUserInfo = async (token: string, info: object) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -335,7 +335,7 @@ export const updateUserInfo = async (token: string, info: object) => {
 
 export const getAndUpdateUserLocation = async (token: string) => {
 	const location = await getUserPosition().catch((err) => {
-		console.log(err);
+		console.error(err);
 		return null;
 	});
 
@@ -343,7 +343,7 @@ export const getAndUpdateUserLocation = async (token: string) => {
 		await updateUserInfo(token, { location: location });
 		return location;
 	} else {
-		console.log('Failed to get user location');
+		console.info('Failed to get user location');
 		return null;
 	}
 };
@@ -363,7 +363,7 @@ export const deleteUserById = async (token: string, userId: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -403,7 +403,7 @@ export const updateUserById = async (token: string, userId: string, user: UserUp
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});

src/lib/apis/utils/index.ts

@@ -15,7 +15,7 @@ export const getGravatarUrl = async (token: string, email: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -41,7 +41,7 @@ export const executeCode = async (token: string, code: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 
 			error = err;
 			if (err.detail) {
@@ -75,7 +75,7 @@ export const formatPythonCode = async (token: string, code: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 
 			error = err;
 			if (err.detail) {
@@ -110,7 +110,7 @@ export const downloadChatAsPDF = async (token: string, title: string, messages:
 			return res.blob();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -136,7 +136,7 @@ export const getHTMLFromMarkdown = async (token: string, md: string) => {
 			return res.json();
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err;
 			return null;
 		});
@@ -170,7 +170,7 @@ export const downloadDatabase = async (token: string) => {
 			window.URL.revokeObjectURL(url);
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});
@@ -206,7 +206,7 @@ export const downloadLiteLLMConfig = async (token: string) => {
 			window.URL.revokeObjectURL(url);
 		})
 		.catch((err) => {
-			console.log(err);
+			console.error(err);
 			error = err.detail;
 			return null;
 		});

src/lib/components/AddConnectionModal.svelte

@@ -30,8 +30,17 @@
 	let url = '';
 	let key = '';
 
+	let connectionType = 'external';
+	let azure = false;
+	$: azure =
+		(url.includes('azure.com') || url.includes('cognitive.microsoft.com')) && !direct
+			? true
+			: false;
+
 	let prefixId = '';
 	let enable = true;
+	let apiVersion = '';
+
 	let tags = [];
 
 	let modelId = '';
@@ -40,7 +49,10 @@
 	let loading = false;
 
 	const verifyOllamaHandler = async () => {
-		const res = await verifyOllamaConnection(localStorage.token, url, key).catch((error) => {
+		const res = await verifyOllamaConnection(localStorage.token, {
+			url,
+			key
+		}).catch((error) => {
 			toast.error(`${error}`);
 		});
 
@@ -50,11 +62,20 @@
 	};
 
 	const verifyOpenAIHandler = async () => {
-		const res = await verifyOpenAIConnection(localStorage.token, url, key, direct).catch(
-			(error) => {
-				toast.error(`${error}`);
-			}
-		);
+		const res = await verifyOpenAIConnection(
+			localStorage.token,
+			{
+				url,
+				key,
+				config: {
+					azure: azure,
+					api_version: apiVersion
+				}
+			},
+			direct
+		).catch((error) => {
+			toast.error(`${error}`);
+		});
 
 		if (res) {
 			toast.success($i18n.t('Server connection verified'));
@@ -85,6 +106,28 @@
 			return;
 		}
 
+		if (azure) {
+			if (!apiVersion) {
+				loading = false;
+
+				toast.error('API Version is required');
+				return;
+			}
+
+			if (!key) {
+				loading = false;
+
+				toast.error('Key is required');
+				return;
+			}
+
+			if (modelIds.length === 0) {
+				loading = false;
+				toast.error('Deployment names are required');
+				return;
+			}
+		}
+
 		// remove trailing slash from url
 		url = url.replace(/\/$/, '');
 
@@ -95,7 +138,9 @@
 				enable: enable,
 				tags: tags,
 				prefix_id: prefixId,
-				model_ids: modelIds
+				model_ids: modelIds,
+				connection_type: connectionType,
+				...(!ollama && azure ? { azure: true, api_version: apiVersion } : {})
 			}
 		};
 
@@ -120,6 +165,14 @@
 			tags = connection.config?.tags ?? [];
 			prefixId = connection.config?.prefix_id ?? '';
 			modelIds = connection.config?.model_ids ?? [];
+
+			if (ollama) {
+				connectionType = connection.config?.connection_type ?? 'local';
+			} else {
+				connectionType = connection.config?.connection_type ?? 'external';
+				azure = connection.config?.azure ?? false;
+				apiVersion = connection.config?.api_version ?? '';
+			}
 		}
 	};
 
@@ -134,7 +187,7 @@
 
 <Modal size="sm" bind:show>
 	<div>
-		<div class=" flex justify-between dark:text-gray-100 px-5 pt-4 pb-2">
+		<div class=" flex justify-between dark:text-gray-100 px-5 pt-4 pb-1.5">
 			<div class=" text-lg font-medium self-center font-primary">
 				{#if edit}
 					{$i18n.t('Edit Connection')}
@@ -171,7 +224,31 @@
 					}}
 				>
 					<div class="px-1">
-						<div class="flex gap-2">
+						{#if !direct}
+							<div class="flex gap-2">
+								<div class="flex w-full justify-between items-center">
+									<div class=" text-xs text-gray-500">{$i18n.t('Connection Type')}</div>
+
+									<div class="">
+										<button
+											on:click={() => {
+												connectionType = connectionType === 'local' ? 'external' : 'local';
+											}}
+											type="button"
+											class=" text-xs text-gray-700 dark:text-gray-300"
+										>
+											{#if connectionType === 'local'}
+												{$i18n.t('Local')}
+											{:else}
+												{$i18n.t('External')}
+											{/if}
+										</button>
+									</div>
+								</div>
+							</div>
+						{/if}
+
+						<div class="flex gap-2 mt-1.5">
 							<div class="flex flex-col w-full">
 								<div class=" mb-0.5 text-xs text-gray-500">{$i18n.t('URL')}</div>
 
@@ -252,6 +329,25 @@
 							</div>
 						</div>
 
+						{#if azure}
+							<div class="flex gap-2 mt-2">
+								<div class="flex flex-col w-full">
+									<div class=" mb-1 text-xs text-gray-500">{$i18n.t('API Version')}</div>
+
+									<div class="flex-1">
+										<input
+											class="w-full text-sm bg-transparent placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-hidden"
+											type="text"
+											bind:value={apiVersion}
+											placeholder={$i18n.t('API Version')}
+											autocomplete="off"
+											required
+										/>
+									</div>
+								</div>
+							</div>
+						{/if}
+
 						<div class="flex gap-2 mt-2">
 							<div class="flex flex-col w-full">
 								<div class=" mb-1.5 text-xs text-gray-500">{$i18n.t('Tags')}</div>
@@ -308,6 +404,11 @@
 										{$i18n.t('Leave empty to include all models from "{{url}}/api/tags" endpoint', {
 											url: url
 										})}
+									{:else if azure}
+										{$i18n.t('Deployment names are required for Azure OpenAI')}
+										<!-- {$i18n.t('Leave empty to include all models from "{{url}}" endpoint', {
+											url: `${url}/openai/deployments`
+										})} -->
 									{:else}
 										{$i18n.t('Leave empty to include all models from "{{url}}/models" endpoint', {
 											url: url
@@ -317,7 +418,7 @@
 							{/if}
 						</div>
 
-						<hr class=" border-gray-100 dark:border-gray-700/10 my-2.5 w-full" />
+						<hr class=" border-gray-100 dark:border-gray-700/10 my-1.5 w-full" />
 
 						<div class="flex items-center">
 							<input

src/lib/components/AddServerModal.svelte

@@ -3,10 +3,6 @@
 	import { getContext, onMount } from 'svelte';
 	const i18n = getContext('i18n');
 
-	import { models } from '$lib/stores';
-	import { verifyOpenAIConnection } from '$lib/apis/openai';
-	import { verifyOllamaConnection } from '$lib/apis/ollama';
-
 	import Modal from '$lib/components/common/Modal.svelte';
 	import Plus from '$lib/components/icons/Plus.svelte';
 	import Minus from '$lib/components/icons/Minus.svelte';
@@ -26,7 +22,6 @@
 	export let edit = false;
 
 	export let direct = false;
-
 	export let connection = null;
 
 	let url = '';
@@ -37,6 +32,9 @@
 
 	let accessControl = {};
 
+	let name = '';
+	let description = '';
+
 	let enable = true;
 
 	let loading = false;
@@ -55,7 +53,7 @@
 		if (direct) {
 			const res = await getToolServerData(
 				auth_type === 'bearer' ? key : localStorage.token,
-				`${url}/${path}`
+				path.includes('://') ? path : `${url}${path.startsWith('/') ? '' : '/'}${path}`
 			).catch((err) => {
 				toast.error($i18n.t('Connection failed'));
 			});
@@ -73,6 +71,10 @@
 				config: {
 					enable: enable,
 					access_control: accessControl
+				},
+				info: {
+					name,
+					description
 				}
 			}).catch((err) => {
 				toast.error($i18n.t('Connection failed'));
@@ -99,6 +101,10 @@
 			config: {
 				enable: enable,
 				access_control: accessControl
+			},
+			info: {
+				name: name,
+				description: description
 			}
 		};
 
@@ -112,6 +118,9 @@
 		key = '';
 		auth_type = 'bearer';
 
+		name = '';
+		description = '';
+
 		enable = true;
 		accessControl = null;
 	};
@@ -124,6 +133,9 @@
 			auth_type = connection?.auth_type ?? 'bearer';
 			key = connection?.key ?? '';
 
+			name = connection.info?.name ?? '';
+			description = connection.info?.description ?? '';
+
 			enable = connection.config?.enable ?? true;
 			accessControl = connection.config?.access_control ?? null;
 		}
@@ -225,12 +237,11 @@
 								</div>
 
 								<div class="flex-1 flex items-center">
-									<div class="text-sm">/</div>
 									<input
 										class="w-full text-sm bg-transparent placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-hidden"
 										type="text"
 										bind:value={path}
-										placeholder={$i18n.t('openapi.json Path')}
+										placeholder={$i18n.t('openapi.json URL or Path')}
 										autocomplete="off"
 										required
 									/>
@@ -240,7 +251,7 @@
 
 						<div class="text-xs text-gray-500 mt-1">
 							{$i18n.t(`WebUI will make requests to "{{url}}"`, {
-								url: `${url}/${path}`
+								url: path.includes('://') ? path : `${url}${path.startsWith('/') ? '' : '/'}${path}`
 							})}
 						</div>
 
@@ -280,6 +291,39 @@
 						{#if !direct}
 							<hr class=" border-gray-100 dark:border-gray-700/10 my-2.5 w-full" />
 
+							<div class="flex gap-2">
+								<div class="flex flex-col w-full">
+									<div class=" mb-0.5 text-xs text-gray-500">{$i18n.t('Name')}</div>
+
+									<div class="flex-1">
+										<input
+											class="w-full text-sm bg-transparent placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-hidden"
+											type="text"
+											bind:value={name}
+											placeholder={$i18n.t('Enter name')}
+											autocomplete="off"
+											required
+										/>
+									</div>
+								</div>
+							</div>
+
+							<div class="flex flex-col w-full mt-2">
+								<div class=" mb-1 text-xs text-gray-500">{$i18n.t('Description')}</div>
+
+								<div class="flex-1">
+									<input
+										class="w-full text-sm bg-transparent placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-hidden"
+										type="text"
+										bind:value={description}
+										placeholder={$i18n.t('Enter description')}
+										autocomplete="off"
+									/>
+								</div>
+							</div>
+
+							<hr class=" border-gray-100 dark:border-gray-700/10 my-2.5 w-full" />
+
 							<div class="my-2 -mx-2">
 								<div class="px-3 py-2 bg-gray-50 dark:bg-gray-950 rounded-lg">
 									<AccessControl bind:accessControl />

src/lib/components/ImportModal.svelte

@@ -0,0 +1,154 @@
+<script lang="ts">
+	import { toast } from 'svelte-sonner';
+	import { getContext, onMount } from 'svelte';
+	const i18n = getContext('i18n');
+
+	import Modal from '$lib/components/common/Modal.svelte';
+	import { extractFrontmatter } from '$lib/utils';
+
+	export let show = false;
+
+	export let onImport = (e) => {};
+	export let onClose = () => {};
+
+	export let loadUrlHandler: Function = () => {};
+	export let successMessage: string = '';
+
+	let loading = false;
+	let url = '';
+
+	const submitHandler = async () => {
+		loading = true;
+
+		if (!url) {
+			toast.error($i18n.t('Please enter a valid URL'));
+			loading = false;
+			return;
+		}
+
+		const res = await loadUrlHandler(url).catch((err) => {
+			toast.error(`${err}`);
+			loading = false;
+			return null;
+		});
+
+		if (res) {
+			if (!successMessage) {
+				successMessage = $i18n.t('Function imported successfully');
+			}
+
+			toast.success(successMessage);
+
+			let func = res;
+			func.id = func.id || func.name.replace(/\s+/g, '_').toLowerCase();
+
+			const frontmatter = extractFrontmatter(res.content); // Ensure frontmatter is extracted
+
+			if (frontmatter?.title) {
+				func.name = frontmatter.title;
+			}
+
+			func.meta = {
+				...(func.meta ?? {}),
+				description: frontmatter?.description ?? func.name
+			};
+
+			onImport(func);
+			show = false;
+		}
+	};
+</script>
+
+<Modal size="sm" bind:show>
+	<div>
+		<div class=" flex justify-between dark:text-gray-300 px-5 pt-4 pb-2">
+			<div class=" text-lg font-medium self-center">{$i18n.t('Import')}</div>
+			<button
+				class="self-center"
+				on:click={() => {
+					show = false;
+				}}
+			>
+				<svg
+					xmlns="http://www.w3.org/2000/svg"
+					viewBox="0 0 20 20"
+					fill="currentColor"
+					class="w-5 h-5"
+				>
+					<path
+						d="M6.28 5.22a.75.75 0 00-1.06 1.06L8.94 10l-3.72 3.72a.75.75 0 101.06 1.06L10 11.06l3.72 3.72a.75.75 0 101.06-1.06L11.06 10l3.72-3.72a.75.75 0 00-1.06-1.06L10 8.94 6.28 5.22z"
+					/>
+				</svg>
+			</button>
+		</div>
+
+		<div class="flex flex-col md:flex-row w-full px-4 pb-3 md:space-x-4 dark:text-gray-200">
+			<div class=" flex flex-col w-full sm:flex-row sm:justify-center sm:space-x-6">
+				<form
+					class="flex flex-col w-full"
+					on:submit|preventDefault={() => {
+						submitHandler();
+					}}
+				>
+					<div class="px-1">
+						<div class="flex flex-col w-full">
+							<div class=" mb-1 text-xs text-gray-500">{$i18n.t('URL')}</div>
+
+							<div class="flex-1">
+								<input
+									class="w-full text-sm bg-transparent disabled:text-gray-500 dark:disabled:text-gray-500 outline-hidden"
+									type="url"
+									bind:value={url}
+									placeholder={$i18n.t('Enter the URL to import')}
+									required
+								/>
+
+								<!-- $i18n.t('Enter the URL of the function to import') -->
+							</div>
+						</div>
+					</div>
+
+					<div class="flex justify-end pt-3 text-sm font-medium">
+						<button
+							class="px-3.5 py-1.5 text-sm font-medium bg-black hover:bg-gray-900 text-white dark:bg-white dark:text-black dark:hover:bg-gray-100 transition rounded-full flex flex-row space-x-1 items-center {loading
+								? ' cursor-not-allowed'
+								: ''}"
+							type="submit"
+							disabled={loading}
+						>
+							{$i18n.t('Import')}
+
+							{#if loading}
+								<div class="ml-2 self-center">
+									<svg
+										class=" w-4 h-4"
+										viewBox="0 0 24 24"
+										fill="currentColor"
+										xmlns="http://www.w3.org/2000/svg"
+										><style>
+											.spinner_ajPY {
+												transform-origin: center;
+												animation: spinner_AtaB 0.75s infinite linear;
+											}
+											@keyframes spinner_AtaB {
+												100% {
+													transform: rotate(360deg);
+												}
+											}
+										</style><path
+											d="M12,1A11,11,0,1,0,23,12,11,11,0,0,0,12,1Zm0,19a8,8,0,1,1,8-8A8,8,0,0,1,12,20Z"
+											opacity=".25"
+										/><path
+											d="M10.14,1.16a11,11,0,0,0-9,8.92A1.59,1.59,0,0,0,2.46,12,1.52,1.52,0,0,0,4.11,10.7a8,8,0,0,1,6.66-6.61A1.42,1.42,0,0,0,12,2.69h0A1.57,1.57,0,0,0,10.14,1.16Z"
+											class="spinner_ajPY"
+										/></svg
+									>
+								</div>
+							{/if}
+						</button>
+					</div>
+				</form>
+			</div>
+		</div>
+	</div>
+</Modal>

src/lib/components/admin/Functions.svelte

@@ -13,6 +13,7 @@
 		exportFunctions,
 		getFunctionById,
 		getFunctions,
+		loadFunctionByUrl,
 		toggleFunctionById,
 		toggleGlobalById
 	} from '$lib/apis/functions';
@@ -32,6 +33,9 @@
 	import Search from '../icons/Search.svelte';
 	import Plus from '../icons/Plus.svelte';
 	import ChevronRight from '../icons/ChevronRight.svelte';
+	import XMark from '../icons/XMark.svelte';
+	import AddFunctionMenu from './Functions/AddFunctionMenu.svelte';
+	import ImportModal from '../ImportModal.svelte';
 
 	const i18n = getContext('i18n');
 
@@ -40,6 +44,8 @@
 	let functionsImportInputElement: HTMLInputElement;
 	let importFiles;
 
+	let showImportModal = false;
+
 	let showConfirm = false;
 	let query = '';
 
@@ -196,6 +202,19 @@
 	</title>
 </svelte:head>
 
+<ImportModal
+	bind:show={showImportModal}
+	loadUrlHandler={async (url) => {
+		return await loadFunctionByUrl(localStorage.token, url);
+	}}
+	onImport={(func) => {
+		sessionStorage.function = JSON.stringify({
+			...func
+		});
+		goto('/admin/functions/create');
+	}}
+/>
+
 <div class="flex flex-col gap-1 mt-1.5 mb-2">
 	<div class="flex justify-between items-center">
 		<div class="flex md:self-center text-xl items-center font-medium px-0.5">
@@ -215,15 +234,36 @@
 				bind:value={query}
 				placeholder={$i18n.t('Search Functions')}
 			/>
+
+			{#if query}
+				<div class="self-center pl-1.5 translate-y-[0.5px] rounded-l-xl bg-transparent">
+					<button
+						class="p-0.5 rounded-full hover:bg-gray-100 dark:hover:bg-gray-900 transition"
+						on:click={() => {
+							query = '';
+						}}
+					>
+						<XMark className="size-3" strokeWidth="2" />
+					</button>
+				</div>
+			{/if}
 		</div>
 
 		<div>
-			<a
-				class=" px-2 py-2 rounded-xl hover:bg-gray-700/10 dark:hover:bg-gray-100/10 dark:text-gray-300 dark:hover:text-white transition font-medium text-sm flex items-center space-x-1"
-				href="/admin/functions/create"
+			<AddFunctionMenu
+				createHandler={() => {
+					goto('/admin/functions/create');
+				}}
+				importFromLinkHandler={() => {
+					showImportModal = true;
+				}}
 			>
-				<Plus className="size-3.5" />
-			</a>
+				<div
+					class=" px-2 py-2 rounded-xl hover:bg-gray-700/10 dark:hover:bg-gray-100/10 dark:text-gray-300 dark:hover:text-white transition font-medium text-sm flex items-center space-x-1"
+				>
+					<Plus className="size-3.5" />
+				</div>
+			</AddFunctionMenu>
 		</div>
 	</div>
 </div>
@@ -447,7 +487,9 @@
 					}
 				}}
 			>
-				<div class=" self-center mr-2 font-medium line-clamp-1">{$i18n.t('Export Functions')}</div>
+				<div class=" self-center mr-2 font-medium line-clamp-1">
+					{$i18n.t('Export Functions')} ({$functions.length})
+				</div>
 
 				<div class=" self-center">
 					<svg

src/lib/components/admin/Functions/AddFunctionMenu.svelte

@@ -0,0 +1,77 @@
+<script lang="ts">
+	import { DropdownMenu } from 'bits-ui';
+	import { flyAndScale } from '$lib/utils/transitions';
+	import { getContext } from 'svelte';
+
+	import Dropdown from '$lib/components/common/Dropdown.svelte';
+	import GarbageBin from '$lib/components/icons/GarbageBin.svelte';
+	import Tooltip from '$lib/components/common/Tooltip.svelte';
+	import Share from '$lib/components/icons/Share.svelte';
+	import DocumentDuplicate from '$lib/components/icons/DocumentDuplicate.svelte';
+	import ArrowDownTray from '$lib/components/icons/ArrowDownTray.svelte';
+	import Switch from '$lib/components/common/Switch.svelte';
+	import GlobeAlt from '$lib/components/icons/GlobeAlt.svelte';
+	import Github from '$lib/components/icons/Github.svelte';
+	import Plus from '$lib/components/icons/Plus.svelte';
+	import Pencil from '$lib/components/icons/Pencil.svelte';
+	import PencilSolid from '$lib/components/icons/PencilSolid.svelte';
+	import Link from '$lib/components/icons/Link.svelte';
+
+	const i18n = getContext('i18n');
+
+	export let createHandler: Function;
+	export let importFromLinkHandler: Function;
+
+	export let onClose: Function = () => {};
+
+	let show = false;
+</script>
+
+<Dropdown
+	bind:show
+	on:change={(e) => {
+		if (e.detail === false) {
+			onClose();
+		}
+	}}
+>
+	<Tooltip content={$i18n.t('Create')}>
+		<slot />
+	</Tooltip>
+
+	<div slot="content">
+		<DropdownMenu.Content
+			class="w-full max-w-[190px] text-sm rounded-xl px-1 py-1.5 z-50 bg-white dark:bg-gray-850 dark:text-white shadow-lg font-primary"
+			sideOffset={-2}
+			side="bottom"
+			align="start"
+			transition={flyAndScale}
+		>
+			<button
+				class="flex rounded-md py-1.5 px-3 w-full hover:bg-gray-50 dark:hover:bg-gray-800 transition"
+				on:click={async () => {
+					createHandler();
+					show = false;
+				}}
+			>
+				<div class=" self-center mr-2">
+					<PencilSolid />
+				</div>
+				<div class=" self-center truncate">{$i18n.t('New Function')}</div>
+			</button>
+
+			<button
+				class="flex rounded-md py-1.5 px-3 w-full hover:bg-gray-50 dark:hover:bg-gray-800 transition"
+				on:click={async () => {
+					importFromLinkHandler();
+					show = false;
+				}}
+			>
+				<div class=" self-center mr-2">
+					<Link />
+				</div>
+				<div class=" self-center truncate">{$i18n.t('Import From Link')}</div>
+			</button>
+		</DropdownMenu.Content>
+	</div>
+</Dropdown>

src/lib/components/admin/Functions/FunctionEditor.svelte

@@ -277,7 +277,7 @@ class Pipe:
 			await tick();
 
 			if (res) {
-				console.log('Code formatted successfully');
+				console.info('Code formatted successfully');
 
 				saveHandler();
 			}

src/lib/components/admin/Settings/Connections/OpenAIConnection.svelte

@@ -62,7 +62,7 @@
 				class="absolute top-0 bottom-0 left-0 right-0 opacity-60 bg-white dark:bg-gray-900 z-10"
 			></div>
 		{/if}
-		<div class="flex w-full">
+		<div class="flex w-full gap-2">
 			<div class="flex-1 relative">
 				<input
 					class=" outline-hidden w-full bg-transparent {pipeline ? 'pr-8' : ''}"

src/lib/components/admin/Settings/Documents.svelte

@@ -91,7 +91,7 @@
 			return;
 		}
 
-		console.log('Update embedding model attempt:', embeddingModel);
+		console.debug('Update embedding model attempt:', embeddingModel);
 
 		updateEmbeddingModelLoading = true;
 		const res = await updateEmbeddingConfig(localStorage.token, {
@@ -114,7 +114,7 @@
 		updateEmbeddingModelLoading = false;
 
 		if (res) {
-			console.log('embeddingModelUpdateHandler:', res);
+			console.debug('embeddingModelUpdateHandler:', res);
 			if (res.status === true) {
 				toast.success($i18n.t('Embedding model set to "{{embedding_model}}"', res), {
 					duration: 1000 * 10
@@ -124,6 +124,13 @@
 	};
 
 	const submitHandler = async () => {
+		if (
+			RAGConfig.CONTENT_EXTRACTION_ENGINE === 'external' &&
+			RAGConfig.EXTERNAL_DOCUMENT_LOADER_URL === ''
+		) {
+			toast.error($i18n.t('External Document Loader URL required.'));
+			return;
+		}
 		if (RAGConfig.CONTENT_EXTRACTION_ENGINE === 'tika' && RAGConfig.TIKA_SERVER_URL === '') {
 			toast.error($i18n.t('Tika Server URL required.'));
 			return;
@@ -143,6 +150,14 @@
 			return;
 		}
 
+		if (
+			RAGConfig.CONTENT_EXTRACTION_ENGINE === 'datalab_marker' &&
+			!RAGConfig.DATALAB_MARKER_API_KEY
+		) {
+			toast.error($i18n.t('Datalab Marker API Key required.'));
+			return;
+		}
+
 		if (
 			RAGConfig.CONTENT_EXTRACTION_ENGINE === 'document_intelligence' &&
 			(RAGConfig.DOCUMENT_INTELLIGENCE_ENDPOINT === '' ||
@@ -163,6 +178,16 @@
 			await embeddingModelUpdateHandler();
 		}
 
+		RAGConfig.ALLOWED_FILE_EXTENSIONS = (RAGConfig?.ALLOWED_FILE_EXTENSIONS ?? '')
+			.split(',')
+			.map((ext) => ext.trim())
+			.filter((ext) => ext !== '');
+
+		RAGConfig.DATALAB_MARKER_LANGS = RAGConfig.DATALAB_MARKER_LANGS.split(',')
+			.map((code) => code.trim())
+			.filter((code) => code !== '')
+			.join(', ');
+
 		const res = await updateRAGConfig(localStorage.token, RAGConfig);
 		dispatch('save');
 	};
@@ -185,7 +210,9 @@
 	onMount(async () => {
 		await setEmbeddingConfig();
 
-		RAGConfig = await getRAGConfig(localStorage.token);
+		const config = await getRAGConfig(localStorage.token);
+		config.ALLOWED_FILE_EXTENSIONS = (config?.ALLOWED_FILE_EXTENSIONS ?? []).join(', ');
+		RAGConfig = config;
 	});
 </script>
 
@@ -246,7 +273,7 @@
 					<hr class=" border-gray-100 dark:border-gray-850 my-2" />
 
 					<div class="mb-2.5 flex flex-col w-full justify-between">
-						<div class="flex w-full justify-between">
+						<div class="flex w-full justify-between mb-1">
 							<div class="self-center text-xs font-medium">
 								{$i18n.t('Content Extraction Engine')}
 							</div>
@@ -256,8 +283,10 @@
 									bind:value={RAGConfig.CONTENT_EXTRACTION_ENGINE}
 								>
 									<option value="">{$i18n.t('Default')}</option>
+									<option value="external">{$i18n.t('External')}</option>
 									<option value="tika">{$i18n.t('Tika')}</option>
 									<option value="docling">{$i18n.t('Docling')}</option>
+									<option value="datalab_marker">{$i18n.t('Datalab Marker API')}</option>
 									<option value="document_intelligence">{$i18n.t('Document Intelligence')}</option>
 									<option value="mistral_ocr">{$i18n.t('Mistral OCR')}</option>
 								</select>
@@ -275,11 +304,156 @@
 									</div>
 								</div>
 							</div>
+						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'datalab_marker'}
+							<div class="my-0.5 flex gap-2 pr-2">
+								<SensitiveInput
+									placeholder={$i18n.t('Enter Datalab Marker API Key')}
+									required={false}
+									bind:value={RAGConfig.DATALAB_MARKER_API_KEY}
+								/>
+							</div>
+
+							<div class="flex justify-between w-full mt-2">
+								<div class="text-xs font-medium">
+									{$i18n.t('Languages')}
+								</div>
+
+								<input
+									class="text-sm bg-transparent outline-hidden"
+									type="text"
+									bind:value={RAGConfig.DATALAB_MARKER_LANGS}
+									placeholder={$i18n.t('e.g.) en,fr,de')}
+								/>
+							</div>
+
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											'Significantly improves accuracy by using an LLM to enhance tables, forms, inline math, and layout detection. Will increase latency. Defaults to True.'
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Use LLM')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_USE_LLM} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t('Skip the cache and re-run the inference. Defaults to False.')}
+										placement="top-start"
+									>
+										{$i18n.t('Skip Cache')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_SKIP_CACHE} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											'Force OCR on all pages of the PDF. This can lead to worse results if you have good text in your PDFs. Defaults to False.'
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Force OCR')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_FORCE_OCR} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											'Whether to paginate the output. Each page will be separated by a horizontal rule and page number. Defaults to False.'
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Paginate')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_PAGINATE} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											'Strip existing OCR text from the PDF and re-run OCR. Ignored if Force OCR is enabled. Defaults to False.'
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Strip Existing OCR')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_STRIP_EXISTING_OCR} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											'Disable image extraction from the PDF. If Use LLM is enabled, images will be automatically captioned. Defaults to False.'
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Disable Image Extraction')}
+									</Tooltip>
+								</div>
+								<div class="flex items-center">
+									<Switch bind:state={RAGConfig.DATALAB_MARKER_DISABLE_IMAGE_EXTRACTION} />
+								</div>
+							</div>
+							<div class="flex justify-between w-full mt-2">
+								<div class="self-center text-xs font-medium">
+									<Tooltip
+										content={$i18n.t(
+											"The output format for the text. Can be 'json', 'markdown', or 'html'. Defaults to 'markdown'."
+										)}
+										placement="top-start"
+									>
+										{$i18n.t('Output Format')}
+									</Tooltip>
+								</div>
+								<div class="">
+									<select
+										class="dark:bg-gray-900 w-fit pr-8 rounded-sm px-2 text-xs bg-transparent outline-hidden text-right"
+										bind:value={RAGConfig.DATALAB_MARKER_OUTPUT_FORMAT}
+									>
+										<option value="markdown">{$i18n.t('Markdown')}</option>
+										<option value="json">{$i18n.t('JSON')}</option>
+										<option value="html">{$i18n.t('HTML')}</option>
+									</select>
+								</div>
+							</div>
+						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'external'}
+							<div class="my-0.5 flex gap-2 pr-2">
+								<input
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
+									placeholder={$i18n.t('Enter External Document Loader URL')}
+									bind:value={RAGConfig.EXTERNAL_DOCUMENT_LOADER_URL}
+								/>
+								<SensitiveInput
+									placeholder={$i18n.t('Enter External Document Loader API Key')}
+									required={false}
+									bind:value={RAGConfig.EXTERNAL_DOCUMENT_LOADER_API_KEY}
+								/>
+							</div>
 						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'tika'}
 							<div class="flex w-full mt-1">
 								<div class="flex-1 mr-2">
 									<input
-										class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+										class="flex-1 w-full text-sm bg-transparent outline-hidden"
 										placeholder={$i18n.t('Enter Tika Server URL')}
 										bind:value={RAGConfig.TIKA_SERVER_URL}
 									/>
@@ -288,27 +462,38 @@
 						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'docling'}
 							<div class="flex w-full mt-1">
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									placeholder={$i18n.t('Enter Docling Server URL')}
 									bind:value={RAGConfig.DOCLING_SERVER_URL}
 								/>
 							</div>
 							<div class="flex w-full mt-2">
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									placeholder={$i18n.t('Enter Docling OCR Engine')}
 									bind:value={RAGConfig.DOCLING_OCR_ENGINE}
 								/>
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									placeholder={$i18n.t('Enter Docling OCR Language(s)')}
 									bind:value={RAGConfig.DOCLING_OCR_LANG}
 								/>
 							</div>
+
+							<div class="flex w-full mt-2">
+								<div class="flex-1 flex justify-between">
+									<div class=" self-center text-xs font-medium">
+										{$i18n.t('Describe Pictures in Documents')}
+									</div>
+									<div class="flex items-center relative">
+										<Switch bind:state={RAGConfig.DOCLING_DO_PICTURE_DESCRIPTION} />
+									</div>
+								</div>
+							</div>
 						{:else if RAGConfig.CONTENT_EXTRACTION_ENGINE === 'document_intelligence'}
 							<div class="my-0.5 flex gap-2 pr-2">
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									placeholder={$i18n.t('Enter Document Intelligence Endpoint')}
 									bind:value={RAGConfig.DOCUMENT_INTELLIGENCE_ENDPOINT}
 								/>
@@ -437,7 +622,7 @@
 							{#if embeddingEngine === 'openai'}
 								<div class="my-0.5 flex gap-2 pr-2">
 									<input
-										class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+										class="flex-1 w-full text-sm bg-transparent outline-hidden"
 										placeholder={$i18n.t('API Base URL')}
 										bind:value={OpenAIUrl}
 										required
@@ -448,7 +633,7 @@
 							{:else if embeddingEngine === 'ollama'}
 								<div class="my-0.5 flex gap-2 pr-2">
 									<input
-										class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+										class="flex-1 w-full text-sm bg-transparent outline-hidden"
 										placeholder={$i18n.t('API Base URL')}
 										bind:value={OllamaUrl}
 										required
@@ -471,7 +656,7 @@
 									<div class="flex w-full">
 										<div class="flex-1 mr-2">
 											<input
-												class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+												class="flex-1 w-full text-sm bg-transparent outline-hidden"
 												bind:value={embeddingModel}
 												placeholder={$i18n.t('Set embedding model')}
 												required
@@ -482,7 +667,7 @@
 									<div class="flex w-full">
 										<div class="flex-1 mr-2">
 											<input
-												class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+												class="flex-1 w-full text-sm bg-transparent outline-hidden"
 												placeholder={$i18n.t('Set embedding model (e.g. {{model}})', {
 													model: embeddingModel.slice(-40)
 												})}
@@ -639,7 +824,7 @@
 									{#if RAGConfig.RAG_RERANKING_ENGINE === 'external'}
 										<div class="my-0.5 flex gap-2 pr-2">
 											<input
-												class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+												class="flex-1 w-full text-sm bg-transparent outline-hidden"
 												placeholder={$i18n.t('API Base URL')}
 												bind:value={RAGConfig.RAG_EXTERNAL_RERANKER_URL}
 												required
@@ -661,7 +846,7 @@
 										<div class="flex w-full">
 											<div class="flex-1 mr-2">
 												<input
-													class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+													class="flex-1 w-full text-sm bg-transparent outline-hidden"
 													placeholder={$i18n.t('Set reranking model (e.g. {{model}})', {
 														model: 'BAAI/bge-reranker-v2-m3'
 													})}
@@ -677,7 +862,7 @@
 								<div class=" self-center text-xs font-medium">{$i18n.t('Top K')}</div>
 								<div class="flex items-center relative">
 									<input
-										class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+										class="flex-1 w-full text-sm bg-transparent outline-hidden"
 										type="number"
 										placeholder={$i18n.t('Enter Top K')}
 										bind:value={RAGConfig.TOP_K}
@@ -692,7 +877,7 @@
 									<div class="self-center text-xs font-medium">{$i18n.t('Top K Reranker')}</div>
 									<div class="flex items-center relative">
 										<input
-											class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+											class="flex-1 w-full text-sm bg-transparent outline-hidden"
 											type="number"
 											placeholder={$i18n.t('Enter Top K Reranker')}
 											bind:value={RAGConfig.TOP_K_RERANKER}
@@ -711,7 +896,7 @@
 										</div>
 										<div class="flex items-center relative">
 											<input
-												class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+												class="flex-1 w-full text-sm bg-transparent outline-hidden"
 												type="number"
 												step="0.01"
 												placeholder={$i18n.t('Enter Score')}
@@ -731,6 +916,26 @@
 									</div>
 								</div>
 							{/if}
+
+							{#if RAGConfig.ENABLE_RAG_HYBRID_SEARCH === true}
+								<div class="mb-2.5 flex w-full justify-between">
+									<div class="self-center text-xs font-medium">
+										{$i18n.t('Weight of BM25 Retrieval')}
+									</div>
+									<div class="flex items-center relative">
+										<input
+											class="flex-1 w-full text-sm bg-transparent outline-hidden"
+											type="number"
+											step="0.01"
+											placeholder={$i18n.t('Enter BM25 Weight')}
+											bind:value={RAGConfig.HYBRID_BM25_WEIGHT}
+											autocomplete="off"
+											min="0.0"
+											max="1.0"
+										/>
+									</div>
+								</div>
+							{/if}
 						{/if}
 
 						<div class="  mb-2.5 flex flex-col w-full justify-between">
@@ -760,6 +965,26 @@
 
 					<hr class=" border-gray-100 dark:border-gray-850 my-2" />
 
+					<div class="  mb-2.5 flex w-full justify-between">
+						<div class=" self-center text-xs font-medium">{$i18n.t('Allowed File Extensions')}</div>
+						<div class="flex items-center relative">
+							<Tooltip
+								content={$i18n.t(
+									'Allowed file extensions for upload. Separate multiple extensions with commas. Leave empty for all file types.'
+								)}
+								placement="top-start"
+							>
+								<input
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
+									type="text"
+									placeholder={$i18n.t('e.g. pdf, docx, txt')}
+									bind:value={RAGConfig.ALLOWED_FILE_EXTENSIONS}
+									autocomplete="off"
+								/>
+							</Tooltip>
+						</div>
+					</div>
+
 					<div class="  mb-2.5 flex w-full justify-between">
 						<div class=" self-center text-xs font-medium">{$i18n.t('Max Upload Size')}</div>
 						<div class="flex items-center relative">
@@ -770,7 +995,7 @@
 								placement="top-start"
 							>
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									type="number"
 									placeholder={$i18n.t('Leave empty for unlimited')}
 									bind:value={RAGConfig.FILE_MAX_SIZE}
@@ -791,7 +1016,7 @@
 								placement="top-start"
 							>
 								<input
-									class="flex-1 w-full rounded-lg text-sm bg-transparent outline-hidden"
+									class="flex-1 w-full text-sm bg-transparent outline-hidden"
 									type="number"
 									placeholder={$i18n.t('Leave empty for unlimited')}
 									bind:value={RAGConfig.FILE_MAX_COUNT}

src/lib/components/admin/Settings/General.svelte

@@ -18,6 +18,7 @@
 	import { compareVersion } from '$lib/utils';
 	import { onMount, getContext } from 'svelte';
 	import { toast } from 'svelte-sonner';
+	import Textarea from '$lib/components/common/Textarea.svelte';
 
 	const i18n = getContext('i18n');
 
@@ -58,10 +59,10 @@
 			};
 		});
 
-		console.log(version);
+		console.info(version);
 
 		updateAvailable = compareVersion(version.latest, version.current);
-		console.log(updateAvailable);
+		console.info(updateAvailable);
 	};
 
 	const updateLdapServerHandler = async () => {
@@ -83,7 +84,7 @@
 		if (res) {
 			saveHandler();
 		} else {
-			toast.error(i18n.t('Failed to update settings'));
+			toast.error($i18n.t('Failed to update settings'));
 		}
 	};
 
@@ -305,6 +306,31 @@
 						<Switch bind:state={adminConfig.SHOW_ADMIN_DETAILS} />
 					</div>
 
+					<div class="mb-2.5">
+						<div class=" self-center text-xs font-medium mb-2">
+							{$i18n.t('Pending User Overlay Title')}
+						</div>
+						<Textarea
+							rows={2}
+							placeholder={$i18n.t(
+								'Enter a title for the pending user info overlay. Leave empty for default.'
+							)}
+							bind:value={adminConfig.PENDING_USER_OVERLAY_TITLE}
+						/>
+					</div>
+
+					<div class="mb-2.5">
+						<div class=" self-center text-xs font-medium mb-2">
+							{$i18n.t('Pending User Overlay Content')}
+						</div>
+						<Textarea
+							placeholder={$i18n.t(
+								'Enter content for the pending user info overlay. Leave empty for default.'
+							)}
+							bind:value={adminConfig.PENDING_USER_OVERLAY_CONTENT}
+						/>
+					</div>
+
 					<div class="mb-2.5 flex w-full justify-between pr-2">
 						<div class=" self-center text-xs font-medium">{$i18n.t('Enable API Key')}</div>
 
@@ -559,6 +585,13 @@
 													/>
 												</div>
 											</div>
+											<div class="flex justify-between items-center text-xs">
+												<div class=" font-medium">Validate certificate</div>
+
+												<div class="mt-1">
+													<Switch bind:state={LDAP_SERVER.validate_cert} />
+												</div>
+											</div>
 											<div class="flex w-full gap-2">
 												<div class="w-full">
 													<div class=" self-center text-xs font-medium min-w-fit mb-1">
@@ -625,6 +658,16 @@
 						<Switch bind:state={adminConfig.ENABLE_USER_WEBHOOKS} />
 					</div>
 
+					<div class="mb-2.5">
+						<div class=" self-center text-xs font-medium mb-2">
+							{$i18n.t('Response Watermark')}
+						</div>
+						<Textarea
+							placeholder={$i18n.t('Enter a watermark for the response. Leave empty for none.')}
+							bind:value={adminConfig.RESPONSE_WATERMARK}
+						/>
+					</div>
+
 					<div class="mb-2.5 w-full justify-between">
 						<div class="flex w-full justify-between">
 							<div class=" self-center text-xs font-medium">{$i18n.t('WebUI URL')}</div>

src/lib/components/admin/Settings/Images.svelte

@@ -198,14 +198,14 @@
 						2
 					);
 				} catch (e) {
-					console.log(e);
+					console.error(e);
 				}
 			}
 
 			requiredWorkflowNodes = requiredWorkflowNodes.map((node) => {
 				const n = config.comfyui.COMFYUI_WORKFLOW_NODES.find((n) => n.type === node.type) ?? node;
 
-				console.log(n);
+				console.debug(n);
 
 				return {
 					type: n.type,

src/lib/components/admin/Settings/Interface.svelte

@@ -1,20 +1,26 @@
 <script lang="ts">
+	import fileSaver from 'file-saver';
+	const { saveAs } = fileSaver;
+
 	import { v4 as uuidv4 } from 'uuid';
 	import { toast } from 'svelte-sonner';
 
-	import { getBackendConfig, getTaskConfig, updateTaskConfig } from '$lib/apis';
+	import { getBackendConfig, getModels, getTaskConfig, updateTaskConfig } from '$lib/apis';
 	import { setDefaultPromptSuggestions } from '$lib/apis/configs';
-	import { config, models, settings, user } from '$lib/stores';
+	import { config, settings, user } from '$lib/stores';
 	import { createEventDispatcher, onMount, getContext } from 'svelte';
 
 	import { banners as _banners } from '$lib/stores';
 	import type { Banner } from '$lib/types';
 
+	import { getBaseModels } from '$lib/apis/models';
 	import { getBanners, setBanners } from '$lib/apis/configs';
 
 	import Tooltip from '$lib/components/common/Tooltip.svelte';
 	import Switch from '$lib/components/common/Switch.svelte';
 	import Textarea from '$lib/components/common/Textarea.svelte';
+	import Spinner from '$lib/components/common/Spinner.svelte';
+	import Banners from './Interface/Banners.svelte';
 
 	const dispatch = createEventDispatcher();
 
@@ -42,6 +48,7 @@
 	const updateInterfaceHandler = async () => {
 		taskConfig = await updateTaskConfig(localStorage.token, taskConfig);
 
+		promptSuggestions = promptSuggestions.filter((p) => p.content !== '');
 		promptSuggestions = await setDefaultPromptSuggestions(localStorage.token, promptSuggestions);
 		await updateBanners();
 
@@ -49,6 +56,7 @@
 	};
 
 	onMount(async () => {
+		await init();
 		taskConfig = await getTaskConfig(localStorage.token);
 
 		promptSuggestions = $config?.default_prompt_suggestions ?? [];
@@ -58,9 +66,40 @@
 	const updateBanners = async () => {
 		_banners.set(await setBanners(localStorage.token, banners));
 	};
+
+	let workspaceModels = null;
+	let baseModels = null;
+
+	let models = null;
+
+	const init = async () => {
+		workspaceModels = await getBaseModels(localStorage.token);
+		baseModels = await getModels(localStorage.token, null, false);
+
+		models = baseModels.map((m) => {
+			const workspaceModel = workspaceModels.find((wm) => wm.id === m.id);
+
+			if (workspaceModel) {
+				return {
+					...m,
+					...workspaceModel
+				};
+			} else {
+				return {
+					...m,
+					id: m.id,
+					name: m.name,
+
+					is_active: true
+				};
+			}
+		});
+
+		console.debug('models', models);
+	};
 </script>
 
-{#if taskConfig}
+{#if models !== null && taskConfig}
 	<form
 		class="flex flex-col h-full justify-between space-y-3 text-sm"
 		on:submit|preventDefault={() => {
@@ -74,8 +113,8 @@
 
 				<hr class=" border-gray-100 dark:border-gray-850 my-2" />
 
-				<div class=" mb-1 font-medium flex items-center">
-					<div class=" text-xs mr-1">{$i18n.t('Set Task Model')}</div>
+				<div class=" mb-2 font-medium flex items-center">
+					<div class=" text-xs mr-1">{$i18n.t('Task Model')}</div>
 					<Tooltip
 						content={$i18n.t(
 							'A task model is used when performing tasks such as generating titles for chats and web search queries'
@@ -100,32 +139,70 @@
 
 				<div class=" mb-2.5 flex w-full gap-2">
 					<div class="flex-1">
-						<div class=" text-xs mb-1">{$i18n.t('Local Models')}</div>
+						<div class=" text-xs mb-1">{$i18n.t('Local Task Model')}</div>
 						<select
 							class="w-full rounded-lg py-2 px-4 text-sm bg-gray-50 dark:text-gray-300 dark:bg-gray-850 outline-hidden"
 							bind:value={taskConfig.TASK_MODEL}
 							placeholder={$i18n.t('Select a model')}
+							on:change={() => {
+								if (taskConfig.TASK_MODEL) {
+									const model = models.find((m) => m.id === taskConfig.TASK_MODEL);
+									if (model) {
+										if (model?.access_control !== null) {
+											toast.error(
+												$i18n.t(
+													'This model is not publicly available. Please select another model.'
+												)
+											);
+										}
+
+										taskConfig.TASK_MODEL = model.id;
+									} else {
+										taskConfig.TASK_MODEL = '';
+									}
+								}
+							}}
 						>
 							<option value="" selected>{$i18n.t('Current Model')}</option>
-							{#each $models.filter((m) => m.owned_by === 'ollama') as model}
+							{#each models as model}
 								<option value={model.id} class="bg-gray-100 dark:bg-gray-700">
 									{model.name}
+									{model?.connection_type === 'local' ? `(${$i18n.t('Local')})` : ''}
 								</option>
 							{/each}
 						</select>
 					</div>
 
 					<div class="flex-1">
-						<div class=" text-xs mb-1">{$i18n.t('External Models')}</div>
+						<div class=" text-xs mb-1">{$i18n.t('External Task Model')}</div>
 						<select
 							class="w-full rounded-lg py-2 px-4 text-sm bg-gray-50 dark:text-gray-300 dark:bg-gray-850 outline-hidden"
 							bind:value={taskConfig.TASK_MODEL_EXTERNAL}
 							placeholder={$i18n.t('Select a model')}
+							on:change={() => {
+								if (taskConfig.TASK_MODEL_EXTERNAL) {
+									const model = models.find((m) => m.id === taskConfig.TASK_MODEL_EXTERNAL);
+									if (model) {
+										if (model?.access_control !== null) {
+											toast.error(
+												$i18n.t(
+													'This model is not publicly available. Please select another model.'
+												)
+											);
+										}
+
+										taskConfig.TASK_MODEL_EXTERNAL = model.id;
+									} else {
+										taskConfig.TASK_MODEL_EXTERNAL = '';
+									}
+								}
+							}}
 						>
 							<option value="" selected>{$i18n.t('Current Model')}</option>
-							{#each $models as model}
+							{#each models as model}
 								<option value={model.id} class="bg-gray-100 dark:bg-gray-700">
 									{model.name}
+									{model?.connection_type === 'local' ? `(${$i18n.t('Local')})` : ''}
 								</option>
 							{/each}
 						</select>
@@ -283,9 +360,9 @@
 
 				<hr class=" border-gray-100 dark:border-gray-850 my-2" />
 
-				<div class="  {banners.length > 0 ? ' mb-3' : ''}">
-					<div class="mb-2.5 flex w-full justify-between">
-						<div class=" self-center text-sm font-semibold">
+				<div class="mb-2.5">
+					<div class="flex w-full justify-between">
+						<div class=" self-center text-sm">
 							{$i18n.t('Banners')}
 						</div>
 
@@ -321,69 +398,13 @@
 						</button>
 					</div>
 
-					<div class=" flex flex-col space-y-1">
-						{#each banners as banner, bannerIdx}
-							<div class=" flex justify-between">
-								<div
-									class="flex flex-row flex-1 border rounded-xl border-gray-100 dark:border-gray-850"
-								>
-									<select
-										class="w-fit capitalize rounded-xl py-2 px-4 text-xs bg-transparent outline-hidden"
-										bind:value={banner.type}
-										required
-									>
-										{#if banner.type == ''}
-											<option value="" selected disabled class="text-gray-900"
-												>{$i18n.t('Type')}</option
-											>
-										{/if}
-										<option value="info" class="text-gray-900">{$i18n.t('Info')}</option>
-										<option value="warning" class="text-gray-900">{$i18n.t('Warning')}</option>
-										<option value="error" class="text-gray-900">{$i18n.t('Error')}</option>
-										<option value="success" class="text-gray-900">{$i18n.t('Success')}</option>
-									</select>
-
-									<input
-										class="pr-5 py-1.5 text-xs w-full bg-transparent outline-hidden"
-										placeholder={$i18n.t('Content')}
-										bind:value={banner.content}
-									/>
-
-									<div class="relative top-1.5 -left-2">
-										<Tooltip content={$i18n.t('Dismissible')} className="flex h-fit items-center">
-											<Switch bind:state={banner.dismissible} />
-										</Tooltip>
-									</div>
-								</div>
-
-								<button
-									class="px-2"
-									type="button"
-									on:click={() => {
-										banners.splice(bannerIdx, 1);
-										banners = banners;
-									}}
-								>
-									<svg
-										xmlns="http://www.w3.org/2000/svg"
-										viewBox="0 0 20 20"
-										fill="currentColor"
-										class="w-4 h-4"
-									>
-										<path
-											d="M6.28 5.22a.75.75 0 00-1.06 1.06L8.94 10l-3.72 3.72a.75.75 0 101.06 1.06L10 11.06l3.72 3.72a.75.75 0 101.06-1.06L11.06 10l3.72-3.72a.75.75 0 00-1.06-1.06L10 8.94 6.28 5.22z"
-										/>
-									</svg>
-								</button>
-							</div>
-						{/each}
-					</div>
+					<Banners bind:banners />
 				</div>
 
 				{#if $user?.role === 'admin'}
 					<div class=" space-y-3">
 						<div class="flex w-full justify-between mb-2">
-							<div class=" self-center text-sm font-semibold">
+							<div class=" self-center text-sm">
 								{$i18n.t('Default Prompt Suggestions')}
 							</div>
 
@@ -466,6 +487,111 @@
 								{$i18n.t('Adjusting these settings will apply changes universally to all users.')}
 							</div>
 						{/if}
+
+						<div class="flex items-center justify-end space-x-2 mt-2">
+							<input
+								id="prompt-suggestions-import-input"
+								type="file"
+								accept=".json"
+								hidden
+								on:change={(e) => {
+									const files = e.target.files;
+									if (!files || files.length === 0) {
+										return;
+									}
+
+									console.log(files);
+
+									let reader = new FileReader();
+									reader.onload = async (event) => {
+										try {
+											let suggestions = JSON.parse(event.target.result);
+
+											suggestions = suggestions.map((s) => {
+												if (typeof s.title === 'string') {
+													s.title = [s.title, ''];
+												} else if (!Array.isArray(s.title)) {
+													s.title = ['', ''];
+												}
+
+												return s;
+											});
+
+											promptSuggestions = [...promptSuggestions, ...suggestions];
+										} catch (error) {
+											toast.error($i18n.t('Invalid JSON file'));
+											return;
+										}
+									};
+
+									reader.readAsText(files[0]);
+
+									e.target.value = ''; // Reset the input value
+								}}
+							/>
+
+							<button
+								class="flex text-xs items-center space-x-1 px-3 py-1.5 rounded-xl bg-gray-50 hover:bg-gray-100 dark:bg-gray-800 dark:hover:bg-gray-700 dark:text-gray-200 transition"
+								type="button"
+								on:click={() => {
+									const input = document.getElementById('prompt-suggestions-import-input');
+									if (input) {
+										input.click();
+									}
+								}}
+							>
+								<div class=" self-center mr-2 font-medium line-clamp-1">
+									{$i18n.t('Import Prompt Suggestions')}
+								</div>
+
+								<div class=" self-center">
+									<svg
+										xmlns="http://www.w3.org/2000/svg"
+										viewBox="0 0 16 16"
+										fill="currentColor"
+										class="w-3.5 h-3.5"
+									>
+										<path
+											fill-rule="evenodd"
+											d="M4 2a1.5 1.5 0 0 0-1.5 1.5v9A1.5 1.5 0 0 0 4 14h8a1.5 1.5 0 0 0 1.5-1.5V6.621a1.5 1.5 0 0 0-.44-1.06L9.94 2.439A1.5 1.5 0 0 0 8.878 2H4Zm4 9.5a.75.75 0 0 1-.75-.75V8.06l-.72.72a.75.75 0 0 1-1.06-1.06l2-2a.75.75 0 0 1 1.06 0l2 2a.75.75 0 1 1-1.06 1.06l-.72-.72v2.69a.75.75 0 0 1-.75.75Z"
+											clip-rule="evenodd"
+										/>
+									</svg>
+								</div>
+							</button>
+
+							{#if promptSuggestions.length}
+								<button
+									class="flex text-xs items-center space-x-1 px-3 py-1.5 rounded-xl bg-gray-50 hover:bg-gray-100 dark:bg-gray-800 dark:hover:bg-gray-700 dark:text-gray-200 transition"
+									type="button"
+									on:click={async () => {
+										let blob = new Blob([JSON.stringify(promptSuggestions)], {
+											type: 'application/json'
+										});
+										saveAs(blob, `prompt-suggestions-export-${Date.now()}.json`);
+									}}
+								>
+									<div class=" self-center mr-2 font-medium line-clamp-1">
+										{$i18n.t('Export Prompt Suggestions')} ({promptSuggestions.length})
+									</div>
+
+									<div class=" self-center">
+										<svg
+											xmlns="http://www.w3.org/2000/svg"
+											viewBox="0 0 16 16"
+											fill="currentColor"
+											class="w-3.5 h-3.5"
+										>
+											<path
+												fill-rule="evenodd"
+												d="M4 2a1.5 1.5 0 0 0-1.5 1.5v9A1.5 1.5 0 0 0 4 14h8a1.5 1.5 0 0 0 1.5-1.5V6.621a1.5 1.5 0 0 0-.44-1.06L9.94 2.439A1.5 1.5 0 0 0 8.878 2H4Zm4 3.5a.75.75 0 0 1 .75.75v2.69l.72-.72a.75.75 0 1 1 1.06 1.06l-2 2a.75.75 0 0 1-1.06 0l-2-2a.75.75 0 0 1 1.06-1.06l.72.72V6.25A.75.75 0 0 1 8 5.5Z"
+												clip-rule="evenodd"
+											/>
+										</svg>
+									</div>
+								</button>
+							{/if}
+						</div>
 					</div>
 				{/if}
 			</div>
@@ -480,4 +606,8 @@
 			</button>
 		</div>
 	</form>
+{:else}
+	<div class=" h-full w-full flex justify-center items-center">
+		<Spinner />
+	</div>
 {/if}

src/lib/components/admin/Settings/Interface/Banners.svelte

@@ -0,0 +1,101 @@
+<script lang="ts">
+	import Switch from '$lib/components/common/Switch.svelte';
+	import Tooltip from '$lib/components/common/Tooltip.svelte';
+	import EllipsisVertical from '$lib/components/icons/EllipsisVertical.svelte';
+	import Sortable from 'sortablejs';
+	import { getContext } from 'svelte';
+	const i18n = getContext('i18n');
+
+	export let banners = [];
+
+	let sortable = null;
+	let bannerListElement = null;
+
+	const positionChangeHandler = () => {
+		const bannerIdOrder = Array.from(bannerListElement.children).map((child) =>
+			child.id.replace('banner-item-', '')
+		);
+
+		// Sort the banners array based on the new order
+		banners = bannerIdOrder.map((id) => {
+			const index = banners.findIndex((banner) => banner.id === id);
+			return banners[index];
+		});
+	};
+
+	$: if (banners) {
+		init();
+	}
+
+	const init = () => {
+		if (sortable) {
+			sortable.destroy();
+		}
+
+		if (bannerListElement) {
+			sortable = Sortable.create(bannerListElement, {
+				animation: 150,
+				handle: '.item-handle',
+				onUpdate: async (event) => {
+					positionChangeHandler();
+				}
+			});
+		}
+	};
+</script>
+
+<div class=" flex flex-col space-y-0.5" bind:this={bannerListElement}>
+	{#each banners as banner, bannerIdx (banner.id)}
+		<div class=" flex justify-between items-center -ml-1" id="banner-item-{banner.id}">
+			<EllipsisVertical className="size-4 cursor-move item-handle" />
+
+			<div class="flex flex-row flex-1 gap-2 items-center">
+				<select
+					class="w-fit capitalize rounded-xl text-xs bg-transparent outline-hidden text-left pl-1 pr-2"
+					bind:value={banner.type}
+					required
+				>
+					{#if banner.type == ''}
+						<option value="" selected disabled class="text-gray-900">{$i18n.t('Type')}</option>
+					{/if}
+					<option value="info" class="text-gray-900">{$i18n.t('Info')}</option>
+					<option value="warning" class="text-gray-900">{$i18n.t('Warning')}</option>
+					<option value="error" class="text-gray-900">{$i18n.t('Error')}</option>
+					<option value="success" class="text-gray-900">{$i18n.t('Success')}</option>
+				</select>
+
+				<input
+					class="pr-5 py-1.5 text-xs w-full bg-transparent outline-hidden"
+					placeholder={$i18n.t('Content')}
+					bind:value={banner.content}
+				/>
+
+				<div class="relative -left-2">
+					<Tooltip content={$i18n.t('Dismissible')} className="flex h-fit items-center">
+						<Switch bind:state={banner.dismissible} />
+					</Tooltip>
+				</div>
+			</div>
+
+			<button
+				class="pr-3"
+				type="button"
+				on:click={() => {
+					banners.splice(bannerIdx, 1);
+					banners = banners;
+				}}
+			>
+				<svg
+					xmlns="http://www.w3.org/2000/svg"
+					viewBox="0 0 20 20"
+					fill="currentColor"
+					class="w-4 h-4"
+				>
+					<path
+						d="M6.28 5.22a.75.75 0 00-1.06 1.06L8.94 10l-3.72 3.72a.75.75 0 101.06 1.06L10 11.06l3.72 3.72a.75.75 0 101.06-1.06L11.06 10l3.72-3.72a.75.75 0 00-1.06-1.06L10 8.94 6.28 5.22z"
+					/>
+				</svg>
+			</button>
+		</div>
+	{/each}
+</div>

src/lib/components/admin/Settings/Models.svelte

@@ -20,6 +20,7 @@
 	import Tooltip from '$lib/components/common/Tooltip.svelte';
 	import Switch from '$lib/components/common/Switch.svelte';
 	import Spinner from '$lib/components/common/Spinner.svelte';
+	import XMark from '$lib/components/icons/XMark.svelte';
 
 	import ModelEditor from '$lib/components/workspace/Models/ModelEditor.svelte';
 	import { toast } from 'svelte-sonner';
@@ -33,6 +34,7 @@
 	import EllipsisHorizontal from '$lib/components/icons/EllipsisHorizontal.svelte';
 	import EyeSlash from '$lib/components/icons/EyeSlash.svelte';
 	import Eye from '$lib/components/icons/Eye.svelte';
+	import { copyToClipboard } from '$lib/utils';
 
 	let shiftKey = false;
 
@@ -166,7 +168,7 @@
 			hidden: !(model?.meta?.hidden ?? false)
 		};
 
-		console.log(model);
+		console.debug(model);
 
 		toast.success(
 			model.meta.hidden
@@ -181,6 +183,17 @@
 		upsertModelHandler(model);
 	};
 
+	const copyLinkHandler = async (model) => {
+		const baseUrl = window.location.origin;
+		const res = await copyToClipboard(`${baseUrl}/?model=${encodeURIComponent(model.id)}`);
+
+		if (res) {
+			toast.success($i18n.t('Copied link to clipboard'));
+		} else {
+			toast.error($i18n.t('Failed to copy link'));
+		}
+	};
+
 	const exportModelHandler = async (model) => {
 		let blob = new Blob([JSON.stringify([model])], {
 			type: 'application/json'
@@ -271,6 +284,18 @@
 						bind:value={searchValue}
 						placeholder={$i18n.t('Search Models')}
 					/>
+					{#if searchValue}
+						<div class="self-center pl-1.5 translate-y-[0.5px] rounded-l-xl bg-transparent">
+							<button
+								class="p-0.5 rounded-full hover:bg-gray-100 dark:hover:bg-gray-900 transition"
+								on:click={() => {
+									searchValue = '';
+								}}
+							>
+								<XMark className="size-3" strokeWidth="2" />
+							</button>
+						</div>
+					{/if}
 				</div>
 			</div>
 		</div>
@@ -381,6 +406,9 @@
 									hideHandler={() => {
 										hideModelHandler(model);
 									}}
+									copyLinkHandler={() => {
+										copyLinkHandler(model);
+									}}
 									onClose={() => {}}
 								>
 									<button
@@ -495,7 +523,7 @@
 						}}
 					>
 						<div class=" self-center mr-2 font-medium line-clamp-1">
-							{$i18n.t('Export Presets')}
+							{$i18n.t('Export Presets')} ({models.length})
 						</div>
 
 						<div class=" self-center">