LibreChat

🎉 LibreChat is joining forces with ClickHouse. Together, we're building the open-source Agentic Data Stack. LibreChat remains 100% open source (MIT license) with the same community-first development approach. Learn more: https://lnkd.in/eJ-jnPUc

LibreChat v0.8.0 is live! Major updates in this release: 🔐 Granular Permissions System – Fine-grained access control with Entra ID integration and ACL-based resource permissions 🏪 Agent Marketplace – Discover and share agents with role-based access controls 🤖 Latest AI Models – GPT-5, Claude Sonnet 4.5, updated DeepSeek & Gemini support ⚙️ Plus: OpenRouter web search, SharePoint integration, MCP improvements, and performance upgrades Thanks to our community and 7 new contributors! Full changelog: https://lnkd.in/eZC7y8Ui

🚀 From Idea to Impact: How LibreChat Became a Strategic AI Platform at Daimler Truck AG In early 2024, we, at Daimler Truck, were gaining momentum in exploring AI - new ideas, different initiatives, and a growing belief in its potential. Yet I found myself reflecting in the evenings about: 💭 How can we be faster? 💭 More scalable? 💭 Closer to the users? 💭 Truly future-proof? Across teams and regions, the patterns were clear: repeated use cases, repetitive architecture, vendor lock-ins, and AI tools built for developers - not for everyone else. We needed a different approach. Not something to buy. Not something to custom-build from scratch. But something in between - something flexible, cost-efficient and reusable. That’s how the idea to integrate LibreChat at Daimler Truck was born. No mandate. No big budget. Just lots of passion and a “can-do” mindset. A Friday evening conversation, a weekend deployment and by Monday, LibreChat was ready to be piloted together with Peter Rothlaender 📈 After alignments with Legal, Worker’s Council, Cybersecurity, Global Communication & HR and a few short weeks, +5,000 colleagues had joined the pilot. It worked. It resonated. It scaled. LibreChat quickly became more than a tool, it became a strategic pillar of our AI enablement and transformation: 🔁 Reusability over repetition 📦 Partnerships and inner-source over vendor lock-in 🚀 Speed over complexity 🧠 Learning by doing 🧍♀️ People before platforms Built on open-source foundations, LibreChat belongs to everyone. Together with partners like Amazon Web Services (AWS) and Slalom, we co-developed several features like: 🔄 Modular Sharepoint connector  🧠 Agent Marketplace 📚 Source citation 🔐 Enterprise granular permissioning All of them contributed back to the open-source community led by Danny Avila - because we believe global partnerships and open-source are fundamental to succeed in the AI era. What began as a passionate initiative is now supported by a small, focused team running & enabling LibreChat globally for Daimler Truck, without turning us into a software company. 💡 Today, all our colleagues around the world can chat with multiple LLMs, build agents, generate images and explore the power of AI - securely and compliantly. And that’s not all, we’re just getting started. LibreChat is now a fundamental pillar of our AI strategy making us one of the first DAX40 companies to adopt an open-source AI platform at scale. Because the real transformation? 👉 It won’t be led by AI. It will be led by people, using AI. Let’s keep building a simpler, faster and stronger Daimler Truck. Let’s get it done. Big thanks to Karlheinz Blank, Patrick Froeger & many others for your continuous support & guidance. It would not have been possible without all of you! #LibreChat #AI #OpenSource #Strategy #Innovation #Leadership #EnterpriseAI #DigitalTransformation #GenerativeAI

“Write a good linked-in post for advertising LibreChat as a great match of Open Source and AI!” – could have been the prompt for any LLM for this post. But I am sure it would not capture what I think why this is a great match. So let me tell as short story: Recently we had our DTA Day in Tokyo, bringing together the Leadership Team of Daimler Truck Asia. Packed with a great agenda about Leadership, our Future and our Strategy. During the Strategy workshop we used the chance to introduce secure ways to use AI with our LibreChat Initiative to the Management Team. I truly believe you should play around a bit, to grasp and learn. The word “grasp” developed from “grab” and similarly in German we have the word “begreifen”. As we run our own LibreChat instance, and we handed out access for everybody to start grasping and experimenting. Naturally I was faced often with the Question “Why can we not use <insert you favorite AI tool here>?”. So why is this so great in comparison with others: ⛑️ Since we run it, it is safe to use with our Data. 🪙 We can run it cost efficient. 🧠 It is multi-Model capable: We can use to any Open LLM and the commercial ones (also means we can just add new capabilities in the same familiar interface). 🌱 Easy access for everyone with single sign-on. 🖖 Multilingual capabilities are crucial in my environment. 🃏 The possibility for everyone to experiment building own Agents or Chatbots with Files. But where it really excels for me as CIO is: LibreChat enables enterprise extensibility. Once somebody has a genius Idea how to apply AI in his area of work, we can do it! We can connect any of our existing enterprise Systems by leveraging APIs, using a (young) standard like MCP or even extend LibreChat if needed. Sure, there is often a lot of Homework to do, like increasing data quality, improving process and the systems itself. But I believe we can make a lot of complex things more accessible with AI. As this post was written without AI, just a reminder: stay critical, value Human connections and don’t lose the ability to think for yourself.    Special thanks to Jordi Higuera and Peter Rothlaender for introducing Librechat to Daimler Truck. They are also actively contributing to LibreChat and strengthening the Open Source ecosystem. https://lnkd.in/ecRH_N7q

Love seeing Opendoor innovate with LibreChat! They're rolling out MCPs and agents to work faster. This is why we build!

LibreChat v0.8.0-rc3 is out now with highly anticipated features! • 🔐 Granular permissions - Share prompts and agents with specific users, groups, roles, or everyone! • 🏪 Agent marketplace - Discover, categorize, and share AI agents • 📁 SharePoint integration - Direct file access with role-based controls • 🛠️ Enterprise Settings - MongoDB pooling, Redis clusters • 🧠 Claude Sonnet 4 - 1M context window (beta) Full changelog: https://lnkd.in/eYM8fAtz

🚀 LibreChat v0.8.0-rc1 is here! Exciting news: we’re moving to a weekly release cycle starting with this version. Check out what’s new and stay tuned for even faster updates! See the v0.8.0-rc1 changelog: 👉 https://lnkd.in/eV7xJ_uM

Network Ninja is hiring a TypeScript Developer to join our AI/ML team - someone who can help maintain and enhance our implementation of LibreChat, an open-source LLM interface. This is a 100% remote position, USA. If you're interested in building in a security-conscious environment, check out the full req. DMs are open, but plz follow instructions: https://lnkd.in/gmDsbU9Y (3x Sales and mini-CFO job posts coming soon.)

AgentHouse is live now and using LibreChat, the ClickHouse MCP server is extremely useful! Try it now to see Agent capabilities (Artifacts, MCP) at work: https://lnkd.in/eSdnnArt

# LibreChat v0.7.7 Released 🚀 Excited to share the latest LibreChat release with significant improvements across the board: ## 🌄 New Models & Reasoning - Added support for Claude 3.7, GPT-4.5, o1-mini, and o3-mini 🌟 - Enhanced reasoning capabilities with thought streaming and strategic function calling 🧠 - Anthropic Agents Reasoning now available across multiple providers ## 🛠️ Expanded Tools & Agent Framework - Improved agent UI with better visualization and artifact management - New tools: OpenWeather integration, YouTube analysis, and Flux image generation 🌦️ - OAuth support for Actions to connect with third-party services - xAI (Grok) Agent now supported with modern interface ## 🔐 Security Upgrades - Two-Factor Authentication with backup codes & QR support - Apple and GitHub Enterprise SSO authentication options - Temporary chat sessions for ephemeral usage ## 💻 Developer Experience - Enhanced artifact editing and downloads for better code workflow - Code Interpreter API now supports Rscript alongside existing languages ## ✨ User Experience - Improved messaging with better editing capabilities - Conversation sharing via URL parameters - Enhanced accessibility and i18Next Multi-language support with 18 languages 🌐 Check out the full changelog: https://lnkd.in/eddasCYg