Skip to main content
FortAPI is pay-as-you-go: your account holds a balance in USD, and each successful call deducts from it based on usage.

​
How charging works

The cost of a request ≈ usage × that model’s price:
  • Input tokens + output tokens: most chat / text models are billed on input and output token counts, and input vs. output usually have different prices.
  • Cache hits: the portion served from prompt cache is usually priced lower.
  • Images / audio: image generation, transcription, TTS, etc. may be billed per image, per second, or on a dedicated token basis.
  • Some models are per-call: a few models have a fixed price per call rather than per-token pricing.
Per-model prices are defined on the main site’s Pricing page. The console Logs page shows exactly how much each request cost.
Balance and rate limits are different things: balance governs “how much total you can spend,” rate limits govern “how many requests per unit time” (see Rate limits).

​
Capping spend per token

When you create a token you can set a spend limit that caps how much that one key can spend. Useful for:
  • Giving a project / script its own budget that stops when exhausted, without affecting other keys on the account;
  • Bounding the blast radius of a key that might leak.
See Authentication & API keys for details.

​
Topping up & checking balance

  • Your current balance is shown on the console home page.
  • Historical usage and per-request detail live under Logs.
  • When the balance runs out, requests return 403 (see Errors); top up to resume.