Models and pricing
Rough cost of a job at each provider. Prices are approximate list prices per million tokens and change often, so check the provider’s own pricing page before relying on them.
| Model | In $/1M | Out $/1M | Your job |
|---|
NVIDIA’s hosted models come with a free tier whose limits vary. Local models (Ollama, LM Studio) cost nothing but your hardware. Prompt caching and batch discounts are not included.