LibreChat
Self-hosted AI chat platform for local models and cloud APIs, with agents, searchable conversations, and multi-user access. MIT licensed.
LibreChat is a self-hosted ChatGPT alternative for people and teams that want local models and cloud AI services in one chat interface. It runs on a server you control and is open source under the MIT license. You can switch models without moving to a separate chat app.
It connects to local backends such as Ollama, Apple MLX, AMD Lemonade, and koboldcpp, as well as OpenAI-compatible APIs. Hosted options include Anthropic's Claude, OpenAI, Google Gemini, AWS Bedrock, and Azure OpenAI. Where model requests go depends on the backend you choose: local models run on your hardware, while hosted providers process requests through their services. Web search and hosted APIs need an internet connection.
LibreChat also supports agents that can work with files, run code, and take API actions. Model Context Protocol (MCP) support lets agents use connected tools and services. Code Interpreter provides isolated code execution, while artifacts let you create and preview React, HTML, and Mermaid content in chat. These features suit teams that want an AI chat workspace with room for more involved tasks.
Conversation search covers messages, files, and code snippets. Memory can carry context between chats. For shared deployments, LibreChat has multi-user authentication with OAuth, SAML, LDAP, and two-factor authentication.
