Who it is forA developer or small team who wants one always-running router instead of a global npm install on each machine.
The problemInstalling and keeping the daemon running on every machine is tedious, and state can be lost when a process restarts.
The outcomeThe daemon runs in a container with state kept in a named volume, and harnesses point at the same endpoint.
What you need first
- npm, to install the package globally
- A terminal on Linux, macOS, or Windows
- A coding harness you already use, such as OpenCode or Claude Code
- Docker, only if you choose the container route
Steps
- Clone the Bansos Router repository and change into its folder. (inferred, check the README)
- Build the image with
docker build -t bansos-router .. - Start the container with
docker run -d --name bansos -p 17070:17070 -v bansos-data:/home/node/.bansos bansos-router. - Check the container logs with
docker compose logs -fif you used compose, or read the container output otherwise. - Point your harnesses at
http://127.0.0.1:17070.
Prompt for your agent
Using the bansos-router repo, build the image with `docker build -t bansos-router .` and run it with the volume and port from the README. Expected output: a running container named bansos. Verify by opening `http://127.0.0.1:17070/healthz` and confirming the healthcheck responds.
Caveat. Inside the container the daemon binds to 0.0.0.0, so publish the port only on a network you trust.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for ihsan-ramadhan/bansos-router and "Host the free model router in a Docker container"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer