# UnoRouter UnoRouter gives developers one OpenAI-compatible API key for models from multiple providers, with smart routing and failover across healthy endpoints. - Findings24: - Website: - Pricing: Freemium - Categories: AI Developer Tools, Developer Tools, LLM API ## What is UnoRouter? UnoRouter is an open source gateway that puts large language models from OpenAI, Anthropic, Google, Meta, DeepSeek, and Qwen behind one OpenAI-compatible API key. Developers can keep their existing SDK and change only the base URL; people using Claude Code, Cline, SillyTavern, or other AI clients can paste one key instead of managing a separate account for each lab. Smart routing sends each call to a healthy provider and retries elsewhere when one fails. The built-in chat stores conversations in the browser rather than on a server. Billing starts with a free tier, followed by pay-as-you-go credit with no subscription. ## What can UnoRouter do? ### Unified OpenAI-compatible API Use one API key and the /v1/chat/completions workflow to connect applications and compatible clients to models from multiple providers. ### Free and paid model access The catalog lists free models marked with a :free suffix alongside paid models billed as you go after a small top-up. ### Client and agent compatibility Setup guides cover coding agents, CLI tools, SillyTavern, RisuAI, and any chat client that accepts a custom OpenAI-compatible API. ### Model comparison Compare available models side by side using factors including price, context length, capabilities, and usage. ### API model authenticity tester Test an endpoint in the browser with deterministic checks for its claimed model identity. Results can show the prompt, raw reply, and matched evidence behind a verdict. ### Multiple API protocols in the tester The testing tool supports OpenAI-compatible chat completions, Anthropic Messages, and Google's native Gemini generateContent format. ## Use Cases ### Build or prototype AI applications Route application requests through a single OpenAI-compatible interface while selecting different model IDs from the UnoRouter catalog. ### Configure coding agents and CLI tools Use UnoRouter as a custom API endpoint for coding environments and terminal tools such as OpenCode, Cline, Roo Code, Aider, and Gemini CLI. ### Compare models before choosing one Review model pricing, context length, capabilities, and usage in side-by-side comparisons to inform model selection. ### Check an unfamiliar API endpoint Run browser-based authenticity checks against an endpoint and inspect the evidence used for the result without sending the API key to UnoRouter's servers. ## Frequently Asked Questions ### What API format does UnoRouter use? Its main routing interface is OpenAI-compatible, including the /v1/chat/completions endpoint. The separate model tester also supports Anthropic Messages and Google's native Gemini API format. ### Are free models available? Yes. Models with a :free suffix cost nothing; they have request limits per account and model. ### How does pricing work? Start on the free tier, then add pay-as-you-go credit that does not expire. No subscription is required. ### Which clients can use UnoRouter? Coding agents, CLI tools, desktop and web chat clients, and any software that accepts a custom OpenAI-compatible API. ### Does the model tester send my API key to UnoRouter? The tester states that tests run in the browser and the key is not sent to UnoRouter's servers. Only the host, model, and score are shared when a user chooses to publish to public rankings. ## Quick Facts - **Category**: AI API routing service - **API style**: OpenAI-compatible chat completions - **Model coverage**: GPT, Claude, Gemini, DeepSeek, GLM, and other model families - **Supported workflows**: Applications, coding agents, CLI tools, and chat clients - **Pricing options**: Free tier, then pay-as-you-go credit - **Website**: unorouter.com