LibreChat
Self-hosted, open-source ChatGPT clone that supports many AI providers.
Quick facts
- Price model
- Self-hosted
- Starting price
- Free (self-hosted)
- Best for
- Self-hosted team chat · Multi-provider AI · Agent experiments
- Replaces
- ChatGPT Team, Enterprise AI portals
- Platforms
- Self-hostedWeb
- Last verified
- 2026-06-22
Why it's listed
Feature-rich open alternative to rented team chatbot dashboards.
LibreChat offers agents, plugins, and multi-user support when self-hosted. Point it at Ollama for free local inference or add API keys for cloud models—your infrastructure, your rules.
The catch
Jump to setup guide ↓Requires server maintenance; feature set can be overwhelming to configure.
How to set up LibreChat
Self-hosted ChatGPT-style UI with agents, plugins, and multi-user support—point at Ollama or BYOK APIs.
- Time
- 45–75 min
- Difficulty
- Technical
- Verified
- 2026-06-22
Before you start
- Docker Desktop or Linux Docker + Compose
- Ollama running if you want free local models (see our Ollama guide)
- MongoDB-compatible stack—official compose bundle includes dependencies
Clone and configure
git clone https://github.com/danny-avila/LibreChat.git && cd LibreChat. Copy .env.example → .env. Generate secrets per docs (CREDS_KEY, JWT secrets). Minimum path: use bundled docker-compose.override.yml for a standard install.
Start the stack
docker compose up -d. First boot pulls several images—allow 5–10 minutes. Web UI defaults to http://localhost:3080.
Register admin user
First signup becomes admin unless registration is disabled in .env. Use a strong password—this is your household AI portal.
Add Ollama endpoint
librechat.yaml → endpoints.custom or ollama section per current docs. Base URL http://host.docker.internal:11434 on Docker Desktop. Select a pulled model in a new chat to verify.
Optional — BYOK providers
Add OpenAI/Anthropic keys in .env or the admin UI. Lets you mix local free chat with paid models for hard tasks—still cheaper than per-seat SaaS.
Optional — invite users
Enable registration or create accounts for family/teammates. Each person gets their own threads on your server.
Troubleshooting
- Blank page on :3080
- docker compose logs api — Mongo not ready yet. Wait and refresh.
- Ollama models missing
- Wrong Docker network URL. From inside LibreChat container, curl the Ollama tags endpoint.
- Too many features enabled
- Start with chat only—disable agents/plugins in config until basics work.
Keep it working
- git pull && docker compose pull monthly—LibreChat moves fast
- Back up Mongo data volume before major upgrades
- For a simpler UI, Open WebUI may be enough (see our Open WebUI guide)
Official docs: www.librechat.ai/docs/quick_start/local
Good fit for
- Dev teams
- Homelab enthusiasts
Not ideal for
- Non-technical solo users
Alternatives
Open WebUI
Self-hosted web interface for chatting with local or BYOK-connected models.
Replaces: ChatGPT Team, Custom GPT enterprise tiers
AnythingLLM
Chat with your documents using local or cloud models—desktop or self-hosted.
Replaces: ChatPDF Pro, Notion AI…
LM Studio
Run local LLMs on your Mac or PC with a friendly desktop app—no API subscription required.
Replaces: ChatGPT Plus, Claude Pro…
Ollama
CLI and API for running open models locally with simple pull-and-run commands.
Replaces: OpenAI API for local tasks, Hosted LLM sandboxes