Every model.
One route.
Discover AI models in one place and explore a single route for inference, with usage denominated in USDC. Choose the right model without managing a separate account with every provider.
Inference without
provider lock-in.
Bring model choice and payment into one workflow. Find the right model first, then route requests through a consistent interface.
Find the right model
Browse reasoning, coding, general, and multimodal models. Compare context windows and indicative token costs.
Route through one API
Switch models by changing the model identifier. The API design follows a familiar chat completions format.
Settle in USDC
A crypto-native balance is designed to pay for metered inference across providers without separate provider billing.
Compare models
Filter the catalog by provider or use case, and inspect cost and context.
Load a balance
Deposit USDC to a wallet-linked account when payments are enabled.
Send requests
Use one API key and model ID to route requests when inference goes live.
Choose by the work.
From fast general models to deeper reasoning, explore options before you integrate.
Rates shown are indicative and may change before live routing.
Find your next model.
Explore the directory, compare costs, and inspect the request format.