Inference providers

NanoGPT auto-routes every request to a strong provider by default, and you can also pin a specific provider per request. Each provider below lists the models it serves, with the exact per-token pricing you pay when selecting it.