ModelRefs / Tokens Per Minute (TPM) — AI Glossary
Tokens Per Minute (TPM) — AI Glossary
An API rate limit specifying the maximum number of tokens (input + output) a caller can consume per minute. Hitting TPM limits returns 429 errors.
Overview
TPM limits prevent overloading provider infrastructure and ensure fair allocation. OpenAI tier-based limits range from 40K TPM (free) to 10M+ TPM (enterprise). Hitting TPM limits returns 429 errors. Mitigation: exponential backoff, rate limiter wrappers (RateLimiterAsync), batching requests, or upgrading tier.
Reference details
| Topic | ecosystem |
|---|---|
| Also known as | TPM, token rate limit |
| Last reviewed | 2026-06-24 |
Related terms
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Tokens Per Minute (TPM) — AI Glossary.
Frequently asked questions
What is Tokens Per Minute (TPM)?
An API rate limit specifying the maximum number of tokens (input + output) a caller can consume per minute.
Is Tokens Per Minute (TPM) the same as TPM?
Yes — TPM, token rate limit are common aliases for Tokens Per Minute (TPM).
What concepts are related to Tokens Per Minute (TPM)?
Closely related concepts include requests per minute, rate limiting, retry with backoff.