One OpenAI-compatible API for 150+ models. Send a prompt and RouteLLM picks the model that answers it best, or call any model by name.
Pay $7 today and get $10 of API credits. Spend them on any model in the catalog.
Get StartedEmail or Google sign-in. Under a minute.
Checkout is the only step between you and the dashboard.
Point your OpenAI SDK at our base URL and ship. Works with Claude Code, Codex CLI and Cursor.
Set model="route-llm" and RouteLLM reads each prompt and sends it to the model that answers it best: frontier models for hard reasoning and code, fast and cheap models for simple tasks. Top-model quality where it matters, lower cost where it does not. Want your own rules? Build a custom router and call it like any other model.
Chat with every model in one workspace, plus the Abacus AI Agent, image and video generation and Abacus AI Desktop.
Sign up and pay $7: you get $10 of API credits to spend on any model.
Right after payment you land on your RouteLLM dashboard. Create an API key or build a custom router, paste the key into your code, and your first call is a minute away.
Yes. Pass any model name from the catalog and the request goes straight to it. Automatic failover between providers keeps your calls running, and prompt caching is preserved.
Each call uses credits in proportion to what the model costs. Closed models are priced at the provider's own rates, never more. Open models match the best price available anywhere. The live per-model prices are in the catalog on the RouteLLM page.
ChatLLM comes with it: chat with every model in one workspace, plus the Abacus AI Agent, image and video generation and Abacus AI Desktop.
We don't use your data for training; it's always encrypted, and we have enterprise compliance, including SOC-2 Type-2 and HIPAA.
Pay $7 to get $10 of API credits
Get Started