Airoute — AI skill for Claude Code
Local AI router.
How to install Airoute
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open foisalislambd/airoute and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Airoute does
Local AI router. One OpenAI-compatible endpoint for every provider you already have, and one key for opencode, claude code, Cursor and any coding agent. Keys stay on your machine.
Alternatives in AI
- Devo — Model-neutral agent desktop/runtime for private, enterprise, and OpenAI-compatible / Anthropic-compatible mode 316 ★
- TurboLLM — Run any local LLM engine, auto-tuned to your GPU — polished web UI + OpenAI/Anthropic-compatible API 251 ★
- Claude Max API Proxy — Actively maintained fork that allows using Claude Code as a model provider via OpenAI compatible protocol 114 ★
README
AIRoute
AIRoute is a local AI router. It runs on your computer, keeps provider keys in a local database, and gives Cursor and other tools one OpenAI-compatible address.
The router listens only on this machine, at `http://127.0.0.1:8787`. Nothing in the app is written to be reached from another computer.
| What you open | Address |
|---|---|
| Web panel | http://127.0.0.1:8787 |
| OpenAI-compatible API | http://127.0.0.1:8787/v1 |
The panel is where you add provider keys, turn models on, and create the router key your tools will use. The API is what Cursor calls.
Longer guides for install, the panel, Cursor, the desktop app, the API, and development are in [docs](docs/README.md).
Install
Node.js 20 or newer is required for the npm package.
npm install -g airoute
airoute
That starts the router and opens the panel in your browser. The same version is also published as `@foisalislambd/airoute` on GitHub Packages.
Desktop builds are attached to each [GitHub release](https://github.com/foisalislambd/airoute/releases). Unzip the archive for your system and run `AIRoute` (`AIRoute.exe` on Windows). The panel folder in that archive is the page the app serves.
Docker uses the same panel and API. The host port stays on this computer:
docker run -d --name airoute \
-p 127.0.0.1:8787:8787 \
-v airoute-data:/data \
foisalislambd/airoute:latest
The image is also `ghcr.io/foisalislambd/airoute`. Details are in [docs/docker.md](docs/docker.md).
First setup
- Open the panel.
- Go to Providers, open a provider, paste that provider's API key, and save it.
- Turn on the models you want to call. A model that is off is not offered to clients.
- Open API keys and create a router key. The secret is shown once and starts with
sk-airoute-. - In Cursor, or any OpenAI client, set the base URL to
http://127.0.0.1:8787/v1and the API key to that router key.
A public model id looks like `openai/gpt-4o-mini`: the provide
Related Skills
Multimodal Vision Skill
One image-recognition engine for any AI agent — 3 protocols (OpenAI Chat/Responses, Anthropic Messages), multi
Multi Model Router
Route AI tasks to the best model using a single OpenAI-compatible API key. This Claude Code skill selects, con
Use Crew Gateway
Route an OpenAI/Anthropic-compatible client through the portable farmd control plane's virtual crew model, for
Local Delegate
MCP that delegates mechanical text tasks to a local OpenAI-compatible LLM endpoint to conserve Claude subscrip
Free LLM Gateway
Local LiteLLM gateway aggregating free LLM providers (Zhipu/SiliconFlow/OpenRouter/NVIDIA/Agnes/SenseNova/dots
Codex Router
本地多模型路由器:让 Codex 桌面端 / Claude Code / 任意 OpenAI 客户端在一个菜单里共用 GPT、DeepSeek、Kimi、GLM、Gemini、Claude、ChatGPT 订阅额度 Lo
Related Agents
Openrouter Agent
Runs a model from any of ~60 vendors through OpenRouter's OpenAI-compatible endpoint, billed per token against
Codex CLI
Codex CLI → any OpenAI-compatible endpoint
Provider Debugger
Diagnose live-provider compatibility failures for Kimi, GLM, and other OpenAI-compatible endpoints