If you're referring to the AI tool, OmniRoute is a free, open-source local AI gateway that sits between your AI applications (OpenCode, Codex CLI, Claude Code, Cursor, Cline, Copilot, etc.) and hundreds of AI providers. It exposes a single OpenAI-compatible API endpoint (typically http://localhost:20128/v1) and automatically routes requests to the best available provider.
Its key features include:
-
One endpoint for everything – configure your coding tools once.
-
Auto-fallback – if one provider reaches its quota or is unavailable, it switches to another automatically.
-
Support for 268+ AI providers, including OpenAI, Anthropic, Google Gemini, Groq, Ollama, LM Studio, DeepSeek, Mistral, NVIDIA NIM, and many others.
-
90+ providers with free tiers, allowing you to maximize free usage.
-
Runs entirely on your machine, so your prompts are sent only to the providers you configure.
-
Token compression to reduce prompt size and lower costs.
Installation is straightforward:
This starts a local web dashboard and API gateway on:
with the OpenAI-compatible endpoint at:
Since you've been using OpenCode, Ollama, Gemini, and local AI models on Gentoo, OmniRoute fits your workflow well. You can:
-
Continue using your local Ollama models.
-
Add Gemini, Claude, GPT, Groq, and other cloud models.
-
Configure OpenCode to point to a single OmniRoute endpoint.
-
Let OmniRoute automatically switch providers when one is slow, unavailable, or out of quota.
It effectively becomes a universal AI backend for all your coding tools.