# Self-hosted alternatives to Onyx

> 2 self-hostable, open-source alternatives to Onyx, ranked by score.

- [private-gpt](https://selfhost.directory/project/private-gpt)
- [Omni](https://selfhost.directory/project/omni)

## Head-to-head

- [private-gpt vs Omni](https://selfhost.directory/compare/omni-vs-private-gpt)

## Also in LLM Integration

Not listed as Onyx alternatives, but from the same category:

- [mcp-searxng](https://selfhost.directory/project/mcp-searxng): Open-source MCP server that adds privacy-focused web search and URL reading to AI clients like Claude and Cursor through a self-hosted SearXNG instance.
- [Manifest](https://selfhost.directory/project/manifest): Self-hosted, open-source LLM gateway: one OpenAI-compatible endpoint routes agents to 300+ models with provider fallbacks, cost tracking and full request logs.
- [oss-llmops-stack](https://selfhost.directory/project/oss-llmops-stack): An open-source, self-hosted LLMOps platform pairing LiteLLM's unified model gateway with Langfuse's tracing, evaluation, and prompt tooling for production AI workloads.
- [DesktopCommanderMCP](https://selfhost.directory/project/desktopcommandermcp): Open-source, self-hosted MCP server that gives Claude, ChatGPT and other AI clients terminal, file and process control on your own machine.
- [9Router](https://selfhost.directory/project/9router): Open-source, self-hosted AI router giving coding CLIs one OpenAI-compatible endpoint, 3-tier fallback across 60+ providers, token compression and a local dashboard.
- [Weave Router](https://selfhost.directory/project/router): Open-source LLM model router for coding agents like Claude Code, Codex, and Cursor; self-hosted with Docker Compose, it picks the cheapest capable model per request.

All 84 LLM Integration projects: https://selfhost.directory/llm-integration

---
Source: https://selfhost.directory/alternative-to/onyx