Overview
Onyx is an open-source AI platform that acts as the application layer for large language models. It gives you a feature-rich chat interface that anyone can host, and it works with all major LLM providers, both self-hosted options like Ollama, LiteLLM, and vLLM, and proprietary ones like Anthropic, OpenAI, and Gemini.
On top of plain chat, Onyx adds advanced capabilities such as retrieval-augmented generation (RAG), web search, code execution, document and image creation, and deep research. It can connect to your own applications through more than 50 built-in indexing connectors or via MCP, so the model can answer questions using your team's real data.
Onyx ships in two editions. The Community Edition is free under the MIT license and covers the core Chat, RAG, Agents, and Actions features. The Enterprise Edition adds features aimed at larger organizations, such as single sign-on, role-based access control, analytics, and whitelabeling.
What it does
- Agentic RAG: hybrid search index combined with AI agents for high-quality information retrieval and answers
- Deep Research: a multi-step research flow that produces in-depth reports
- Custom Agents: build AI agents with their own instructions, knowledge, and actions
- Web Search: browse the web for up-to-date information, with support for Serper, Google PSE, Brave, SearXNG, Firecrawl, Exa, and a built-in crawler
- Actions & MCP: let agents interact with external applications using flexible auth options
- Code Execution, Artifacts, Voice Mode, and Image Generation for analyzing data, generating downloadable files, and multimodal chat
Getting started
The fastest way to try Onyx is the one-line install script, which sets up a Docker-based deployment for you. You can also try it without deploying anything by using the hosted Onyx Cloud.
Install Onyx with a single command
Run the official install script. It deploys Onyx using Docker on your machine.
curl -fsSL https://onyx.app/install_onyx.sh | bashChoose a deployment mode
Onyx Lite is a lightweight Chat UI that runs in under 1GB of memory, ideal for quick tests or teams that only want chat and agents. Standard Onyx adds the vector + keyword RAG index, background workers for connector syncing, model inference servers, and performance components like Redis and MinIO. Detailed deployment guides for Docker, Kubernetes, and Helm/Terraform are in the docs.
Connect your LLM and data
Point Onyx at any major LLM provider, self-hosted (Ollama, LiteLLM, vLLM) or proprietary (Anthropic, OpenAI, Gemini). Then add data using the 50+ built-in indexing connectors or MCP so agents can search your own content.
Commands and code are distilled from the project's own documentation — always check the official repo for the latest.
When to use it
- Self-host an internal AI chat assistant that answers questions grounded in your company's documents and applications via RAG
- Run deep research workflows that gather information across multiple steps and produce detailed reports
- Build custom AI agents with their own instructions, knowledge, and actions that can call external tools through MCP
- Give a team a private, provider-agnostic chat UI that works with both self-hosted and proprietary LLMs
How Onyx compares
Onyx alongside other open-source chat uis tools AI/TLDR tracks, ranked by GitHub stars.
| Tool | Stars | What it does |
|---|---|---|
| Open WebUI | ★ 142k | A self-hosted web interface for chatting with local and API-based LLMs (Ollama, OpenAI-compatible) that adds RAG, web search, image generation, and multi-user access. |
| Lobe Chat | ★ 78.9k | An open-source chat UI with a plugin and agent marketplace, multi-model support, voice, vision, and one-click self-hosting. |
| AnythingLLM | ★ 61.8k | A desktop and self-hosted chat app focused on document Q&A and agents, letting you build private RAG workspaces over your own files and data sources. |
| Cherry Studio | ★ 47.5k | A cross-platform desktop chat client that connects to many cloud and local model providers and ships with hundreds of preset assistants and MCP support. |
| Chatbox | ★ 40.6k | A desktop and web chat client for ChatGPT, Claude, Gemini, Ollama, and other models, with local message storage and multi-provider switching. |
| LibreChat | ★ 39.5k | A self-hostable ChatGPT-style chat app that unifies many AI providers and adds agents, MCP support, artifacts, a code interpreter, and multi-user authentication. |
| Onyx | ★ 30.4k | Open-source AI chat platform with RAG, web search, agents, and 50+ connectors |
| Vercel AI Chatbot | ★ 20.5k | An open-source Next.js starter template for building a full-featured AI chatbot with the Vercel AI SDK, auth, and multi-model routing. |