Gateway · Models
Every model Gateway carries, with prices by prompt size. Expand a model for every tier, cache charges and prices by setting. A second, lower price is for a provider that keeps data.
28 live
Your prompt size chooses the tier. It includes new input, cached input and input saved to cache. That tier’s rates apply to the request.
When a provider changes a price, this page changes with it. A model appears here only once it has one.
Input and output are counted separately and drawn from your prepaid balance as each reply finishes.
Choose one model for chat and another for automations or background work, right in the app.
Your Otto takes the cheaper provider. Choose Zero Data Retention in the app and providers keep nothing.