UnoRouter

What is UnoRouter?

UnoRouter is an LLM gateway for engineering teams that collapses many AI vendor integrations into one key and one prepaid balance, with failover handled at the gateway.

Instead of provisioning a key, a quota and an invoice with each model vendor, your services call a single OpenAI-compatible endpoint. Model choice becomes a request parameter, so a team can test a new model in staging and roll it out by changing one value.

Reliability lives in the gateway rather than in every client. Each model can be served through more than one upstream route, and when the active route errors, slows down or returns rate limits, requests shift to a healthy one without retries in your code.

Costs stay simple: free models and a free tier for experiments, then pay as you go against one balance, with no subscription. The project is open source end to end, works with Claude Code, Cline and OpenCode as well as the OpenAI SDKs, and its web chat keeps conversations in the browser.

UnoRouter features

• One OpenAI-compatible key and one prepaid balance for many model vendors
• Automatic failover to a healthy route when a provider errors or rate limits
• Works with the OpenAI SDKs, Claude Code, Cline and OpenCode

Who uses UnoRouter and for what

• Teams that want to swap or test models without new vendor contracts
• Coding agent users who switch models per task from one base URL

UnoRouter FAQ

Comments (0)

No comments yet. Ask the maker a question or share how you would use it.