Simple, call-based pricing
Gateway pricing covers governance and logging. You pay your model provider separately for tokens.
Questions
- What counts as a call?
- Each successful request through /v1/chat. Rejected and failed requests do not count.
- What happens at the limit?
- Further calls return a 429 quota_exceeded error until the month resets or you upgrade.
- Do you store my prompts?
- No. Gateway stores a hash of each prompt, with model, tokens, latency, and decision.