LibreChat
LibreChat is a self-hosted AI chat platform under MIT that unifies the major AI providers in one interface: Anthropic, OpenAI, Google, AWS Bedrock, Azure, Vertex AI, Ollama and any OpenAI-compatible API. Beyond chat it has agents with MCP servers and tools, a code sandbox for eight languages, web search, file chat, image generation, artifacts, and multi-user sign-in with OAuth, LDAP or email. It runs with Docker.
What does LibreChat do?
LibreChat is "a self-hosted AI chat platform that unifies all major AI providers in a single, privacy-focused interface." The README lists Anthropic, AWS Bedrock, OpenAI, Azure OpenAI, Google, Vertex AI and the OpenAI Responses API, and any OpenAI-compatible API as a custom endpoint "no proxy required", with Ollama, Groq, Mistral, OpenRouter, DeepSeek and others named. You can branch and fork conversations, and import chats from ChatGPT.
Its agents use MCP servers, tools, file search and code execution, with skills, subagents and an agent marketplace. The code sandbox runs Python, Node.js, Go, C and C++, Java, PHP, Rust and Fortran. Web search combines search providers, scrapers and re-rankers. Images come from GPT-Image-1, DALL-E, Stable Diffusion, Flux or any MCP server. Artifacts render React, HTML and Mermaid in the chat. Speech runs through OpenAI, Azure or ElevenLabs. For teams it has OAuth2, LDAP and email sign-in, moderation, token-spend tools and an admin panel for users, groups and roles. The interface lists 32 languages, and responses resume after a dropped connection.
Key facts
- Licence: MIT, copyright LibreChat, with no added conditions.
- Providers: Anthropic, AWS Bedrock, OpenAI, Azure OpenAI, Google, Vertex AI, OpenAI Responses API, and any OpenAI-compatible endpoint including Ollama, Groq, Mistral, OpenRouter and DeepSeek.
- Agents with MCP servers, tools, file search and code execution; skills, subagents and an agent marketplace.
- Code sandbox for Python, Node.js, Go, C and C++, Java, PHP, Rust and Fortran.
- Web search, file chat, image generation (GPT-Image-1, DALL-E, Stable Diffusion, Flux), artifacts, and speech through OpenAI, Azure or ElevenLabs.
- Multi-user: OAuth2, LDAP and email sign-in, moderation, token-spend controls, and an admin panel for users, groups and roles.
- Runs with Docker and Docker Compose; one-click templates exist for Railway, Zeabur and Sealos, which are third-party hosts.
- Every GitHub release is marked pre-release, so the API reports no latest release; the newest tag when read was v0.8.8-rc3 on 2026-09-14. 44,247 stars on 2026-09-17.
What does it replace, and where does it fall short?
LibreChat replaces ChatGPT and Claude for a team that wants both, and every other provider, behind one self-hosted interface with agents, a code sandbox and sign-in, under a plain MIT licence it can rebrand. It is the pick for Anthropic built in on the open-source alternatives to Claude page.
Where it falls short: it is a server to run, not an app to install, and it has no desktop or mobile client of its own. Several features are marked experimental or beta in the README, and every GitHub release is tagged pre-release. For a single person on a laptop, Jan or Cherry Studio is simpler.
How does LibreChat run?
Run it with Docker Compose; the README says its stacks give one-command setup, and it works fully local or deployed to the cloud. Add provider keys for the hosted models you want, or point a custom endpoint at Ollama or another OpenAI-compatible server. Document search uses a separate companion service, rag_api. The README names no paid tier from the maintainer; the Railway, Zeabur and Sealos buttons deploy it to those hosts.
Who is LibreChat for?
A team or a company that wants one private chat front end for every model it pays for, with agents, code execution and sign-in, under a licence it can brand as its own. One person who wants a desktop app should look at Jan, Chatbox or Cherry Studio.
What limits does the README state?
From the README: attached workspaces are "highly experimental", the agent management API is in beta, agent plugins are experimental, and "Please consult the changelog for breaking changes before updating."
Questions people ask
Is LibreChat open source?
Yes. LibreChat is MIT licensed with no added conditions, so you may use, modify, rebrand and sell services built on it. The README names no paid tier from the maintainer; the one-click deploy buttons put it on third-party hosts.
Does LibreChat work with Claude?
Yes. Anthropic is the first provider listed in the README, alongside AWS Bedrock, OpenAI, Azure OpenAI, Google and Vertex AI. Any OpenAI-compatible API also works as a custom endpoint, which covers Ollama and most local servers.
Can LibreChat run code like ChatGPT?
Yes. Its code sandbox runs Python, Node.js, Go, C and C++, Java, PHP, Rust and Fortran, and its agents can call code execution alongside MCP servers, tools and file search. Artifacts render React, HTML and Mermaid inside the chat.
How does LibreChat compare with ChatGPT and Claude?
It gives a team both, and every other provider, in one self-hosted interface with agents, code execution, web search and sign-in, under MIT. ChatGPT and Claude are each one vendor's hosted app. The open-source alternatives to Claude page compares LibreChat with Open WebUI, AnythingLLM and the desktop clients.
Why does GitHub show no latest release for LibreChat?
Every release in the repository is marked as a pre-release, so GitHub's latest-release endpoint returns nothing. When read, the newest tag was v0.8.8-rc3 on 2026-09-14, and the newest without rc in its name was v0.8.7 in June, also marked pre-release. The facts panel shows what the API returns.
Sources
- LibreChat README and MIT LICENSE: github.com/danny-avila/LibreChat, read 2026-09-17.
Compared with the others
On the open-source alternatives to ChatGPT page, LibreChat is the pick for every provider in one mit platform. Anthropic, OpenAI, Google, Bedrock, Azure and Ollama in one interface, with agents, MCP, a code sandbox for eight languages, web search, artifacts and sign-in.
Also on that page: Open WebUI for a private chatgpt for a whole team, AnythingLLM for chatting with your documents, Jan for a private chatgpt on one laptop.
On the open-source alternatives to Claude page, LibreChat is the pick for claude plus agents, artifacts and a code sandbox. Anthropic is first in its provider list, alongside Bedrock and Vertex, with agents, MCP, artifacts, a code sandbox for eight languages and sign-in, under MIT.
Also on that page: Cherry Studio for many providers and 300+ assistants in one desktop app, Chatbox for one app across desktop and phone, AnythingLLM for chatting with your documents.
More chat interfaces
A self-hosted chat platform for Ollama and OpenAI-compatible models with document search over 9 vector databases, web search, image generation, roles and single sign-on, under a BSD-3 licence with an added branding clause.
LobeHub, formerly LobeChat, is an agent workspace with an agent builder, agent groups, scheduled runs, editable memory and a library of over 10,000 tools and MCP plugins, under a community licence that needs a paid licence to distribute a modified version.
An MIT all-in-one AI app from Mintplex Labs for chatting with your documents with citations, running agents with MCP, and using any of 39 model providers including local llama.cpp and Ollama, as a desktop app or a multi-user Docker server.
An AGPL-3.0 desktop client for Windows, macOS and Linux that connects OpenAI, Gemini, Anthropic and local Ollama or LM Studio models, with 300+ preset assistants, side-by-side model chats, file support and MCP servers.
An AGPL-3.0 desktop app for local models, formerly text-generation-webui, with five engines including llama.cpp and ExLlamaV3, an OpenAI and Anthropic-compatible API, tool calling, file chat, image generation and LoRA training, fully offline.
An Apache-2.0 desktop app from Menlo Research that downloads and runs open models from Hugging Face on your own computer, connects to OpenAI, Anthropic and others when you want, serves a local OpenAI-compatible API, and supports MCP.
Added September 17, 2026. Every claim above comes from the project's README, LICENSE or model card, read on September 17, 2026, or from the GitHub API on the date shown in the panel. Found an error? Write to reveneau@licheo.com and it is fixed in the next weekly pass. Repository: github.com/danny-avila/LibreChat.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.