LibreChat
Self-hosted, multi-provider AI chat interface with agents, MCP, and search.
Integrate Claude, OpenAI, and local models into a secure, feature-rich chat environment.
LibreChat is an open-source, customizable ChatGPT clone designed to unify multiple AI providers under a single self-hosted interface. It suits mental health professionals and clinics wanting complete control over their client interactions and clinical notes across various models.
What it is
The platform centralizes connections to Anthropic, OpenAI, AWS Bedrock, Google, and local endpoints while adding agent creation, Model Context Protocol (MCP) support, and file uploads. It includes advanced multi-user authentication, admin panels, and tracing tools to inspect conversations step by step.
Who it's for
It suits tech-savvy mental health practitioners and clinic administrators who want to deploy private AI solutions with strict user controls. It does not suit clinicians looking for turnkey, HIPAA-certified cloud EHR apps without a self-hosting setup.
Key features
Model Selection
Switch between Anthropic, OpenAI, AWS, Google, and local providers mid-chat.
Agents & Tools
Build custom assistants using Model Context Protocol and file search.
Sandboxed Execution
Run code securely in isolated environments across multiple programming languages.
Resumable Streams
Never lose responses with auto-reconnecting streams and multi-tab synchronization.
Admin & Auth
Manage multi-user access securely with OAuth2, LDAP, and role-based permissions.
Up and running in 3 steps
Our honest verdict
What's great
- Supports a massive range of commercial and local AI endpoints without needing proxies.
- Provides granular multi-user authentication and a dedicated admin panel for role management.
- Features built-in Model Context Protocol (MCP) support for extensible tool integration.
- Offers resumable streams that sync chats across multiple devices and tabs.
Where it falls short
- Requires managing your own self-hosted infrastructure, Docker containers, and environment variables.
- API usage costs for commercial models like Claude or GPT-4 apply independently of the software.
- Advanced features like attached workspaces and trace viewers are experimental and prone to change.
Pricing & license
Free under the MIT license. The only cost is running the self-hosted infrastructure and paying for any commercial AI model APIs you choose to integrate.
Open remote AI-training roles
876 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Can I run LibreChat entirely offline?
Yes, by connecting local model providers like Ollama or koboldcpp through custom endpoints, you can operate without sending data to external APIs.
How does multi-user authentication work?
LibreChat supports secure email login, OAuth2, and LDAP integration, paired with an admin panel to assign specific roles and group permissions.
