Models

Small open-weight models available through the API.

ModelIdContextPrecisionInput / MtokCache read / MtokCache write / MtokOutput / MtokStatus
Qwen3 0.6Btokencannon/qwen3-0.6b32Kbfloat16$0.05$0.005$0.00$0.15active
Qwen3 1.7Btokencannon/qwen3-1.7b32Kbfloat16$0.10$0.01$0.00$0.25active

Prices are per million tokens across four billed dimensions: input, cache read, cache write, and output. They are read live from GET /v1/models, which returns the catalog in OpenAI format. Model reference.

Want another model supported? Email Nick.