What are the Model API rate limits?
Meta Model API enforces rate limits to ensure fair usage across all customers. Limits apply per team and are measured in two dimensions:
- Requests per minute (RPM) — the number of API calls you can make per minute.
- Tokens per minute (TPM) — the total tokens (input + output) processed per minute.
You can view your team's per-model limits on the Limits page — a table of the requests per minute and tokens per minute allowed for each model.
What happens when you exceed a limit
When you exceed a rate limit, the API returns an HTTP 429 error ("too many requests"). Requests resume automatically once your usage drops below the threshold.
Best practices
- Know your limits. Check the Limits page to see the RPM and TPM allowed for each model on your team.
- Optimize your prompts. Remove unnecessary words and instructions to reduce token consumption.
- Implement exponential backoff. When you receive a 429 error, wait before retrying. Increase the wait time with each successive retry rather than retrying immediately.
- Monitor your usage. Track your consumption on the Usage page to anticipate and prevent rate-limit issues.
For per-model limits, retry strategies, and how to request higher limits, see Rate limits in the developer documentation.
If rate-limit issues persist, contact support.