AI API Rate Limit Capacity Calculator (TPM & RPM → requests per day)

How it works

Requests per minute allowed by the token limit = TPM ÷ tokens per request; the real ceiling is the smaller of that and the RPM limit, times your safety margin. Your daily capacity is set by the busiest hour: busiest-hour requests = usable per minute × 60, and daily requests = that ÷ the busiest hour’s share of the day. Providers may count input and output tokens differently against limits and raise limits with usage tiers, so check your account’s limits page.

⚠ Estimates for information only — not financial advice.

FAQ

What happens when I hit the limit?

The API returns an error (often HTTP 429) and you must retry later. Use retries with exponential back-off and queue non-urgent work.

How do I raise my limits?

Most providers raise limits automatically as you spend more (usage tiers) or on request. Batch APIs usually have separate, larger limits.

Checked 10 Nov 2026. Rate-limit arithmetic for per-minute token (TPM) and request (RPM) limits as used by the major AI API providers; the limits are the ones shown for your own account.