Skip to content

v0.8.1

Pre-release
Pre-release

Choose a tag to compare

View all tags
@danny-avila danny-avila released this 11 Dec 19:52
· 2916 commits to main since this release
v0.8.1
b288d81
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

🏞️ Highlights

Major features and improvements since v0.8.0

🤖 Expanded Model Support

Support for the latest AI models including Gemini 3, GPT-5.1, Grok 4, Claude Opus/Haiku 4.5, and Zhipu AI GLM family, with comprehensive pricing and context window configurations.

📎 Upload to Provider Support

Files can now be directly uploaded to and handled by supported providers, removing the need for intermediate OCR and other multimodal processing steps.

🖼️ Improved Artifact Management and Shareability

Artifacts have been improved with version control capabilities, better UI interactions and rendering support, and can now be viewed within shared conversations.

🤖 Agent Handoffs (Beta)

Agents can now be configured to transfer control of conversations to other specialist agents when specific expertise is needed.

⚙️ Other Notable Changes

  • Langfuse Tracing support
  • Helicone AI Gateway integration
  • Bedrock Prompt Caching support
  • Screens are now prevented from sleeping during response generation
  • File names can now be used as context for Anthropic document uploads
  • Distributed Leader Election with Redis for multi-server-instance coordination within cluster deployments
  • Whether specific tool badges (file search, mcp servers, etc.) will be rendered in the chat input can now be configured via Model Specs
  • Accessibility improvements across the entire app for better keyboard navigation, screen reader support, and accessibility standards compliance
  • Helm Chart enhancements: custom deployment labels, annotations, init containers, ConfigMaps support, and additional RAG API environment variables
  • MCP Registry System restructure with distributed caching, leader-follower coordination, timeout-protected server initialization, and user-scoped tool caching

https://www.librechat.ai/changelog/v0.8.1

Click to expand Release Notes

✨ Features

🐛 Bug Fixes

📦 Dependencies & Chores

🔧 Refactoring

  • 🔄 refactor: Migrate Cache Logic to TypeScript by @nhtruong in #9771
  • 🫴 refactor: Broader Support for GPT-OSS Naming by @danny-avila in #9978
  • ⚡ refactor: Latest Message Tracking with Robust Text Key Generation by @danny-avila in #10059
  • 📡 refactor: Flush Redis Cache Script by @danny-avila in #10087
  • 🪂 refactor: OCR Fallback for "Upload as Text" File Process by @danny-avila in #10126
  • ♻️ refactor: Replace fontSize Recoil atom with Jotai by @berry-13 in #10171
  • 🔄 refactor: MCP Registry System with Distributed Caching by @nhtruong in #10191
  • ♻️ refactor: Message Cache Clearing Logic into Reusable Helper by @danny-avila in #10226
  • 🎪 refactor: Allow Last Model Spec Selection without Prioritizing by @danny-avila in #10258
  • 🔄 refactor: OAI Image Edit Proxy, Speech Settings Handling, Import Query Data Usage by @danny-avila in #10281
  • 📫 refactor: OpenID Email Claim Fallback by @danny-avila in #10296
  • 🔄 refactor: Max tokens handling in Agent Initialization by @danny-avila in #10299
  • 🧠 refactor: Improve Reasoning Component Structure and UX by @danny-avila in #10320
  • ⚗️ refactor: Provider File Validation with Configurable Size Limits by @danny-avila in #10405
  • 🤖 refactor: Improve Agent Handoff Context Tracking by @danny-avila in #10553
  • 🔧 refactor: Update Avatar component to improve file selection handling by @usnavy13 in #10555
  • 🔬 refactor: Prevent Automatic MCP Server UI Deselection by @danny-avila in #10588
  • 📑 refactor: Skip H1 Rendering for Falsy Header Values in AuthLayout by @NalinNair in #10606
  • 🪂 refactor: MCP Server Init Fallback by @danny-avila in #10608
  • 📄 refactor: Add Provider Fallback for Media Encoding using Client Endpoint by @ZihaoZhou in #10656
  • 🪙 refactor: Remove Use of CHECK_BALANCE in Balance Scripts by @ineiti in #10702
  • ⬇️ refactor: Improve Conversation JSON Export by @danny-avila in #10726
  • 📬 refactor: Normalize Email Handling in User Methods by @danny-avila in #10743
  • 📐 refactor: Exclude Params from OAI Reasoning Models by @danny-avila in #10745
  • 🧊 refactor: iconURL Handling in Submission Flow by @danny-avila in #10772
  • ⚡ refactor: Optimize & Standardize Tokenizer Usage by @danny-avila in #10777
  • 🪟 refactor: Portal Control for Shared View Settings and Google Balance Support by @danny-avila in #10786
  • 📦 refactor: Request Message Sanitization for Smaller Final Response by @danny-avila in #10792
  • 📑 refactor: File Search Citations Dual-Format Unicode Handling by @danny-avila in #10888
  • 📂 refactor: File Type Inference for Frontend File Validation by @danny-avila in #10807
  • 🖇️ refactor: Improve prompt for Better Citation Formatting by @mludvig in #10858

🌍 Internationalization

📚 Documentation

🎨 UI/UX

New Contributors

Full Changelog: v0.8.0...v0.8.1

Assets 2
Loading