LLM Router - Open-Weight Models x402 REST usage
An LLM router for an open-weight inference service. It analyzes your request, selects a suitable compatible model, runs the selected model, and returns its answer in OpenAI-compatible format.
0.004748 USDC per call
per call