How billing works

Charged on actual token usage; how failed requests are handled.

Last updated 2026-08-17

Charged on real usage

Chat models are billed on input + output tokens at the prices shown on the Models page. Cache hits, reasoning tokens, audio and images are priced separately and itemised on your bill.

Failed requests

• Our gateway failed: not charged
• Upstream rejected your parameters: not charged
• Upstream had already produced output before the failure: charged for that part

The test is whether upstream actually spent compute, not whether you got a 200.

• Next: Top-up methods and crediting time

Help center