Rate Limit (TPM/RPM) Planner
Check if your traffic fits provider TPM and RPM limits.
With the default inputs, Within limits — tokens-per-minute (TPM) is tightest — TPM 50.0% · RPM 5.0%. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.
Results update automatically as you type.
- Tokens / minute used
- 1,000,000
- TPM limit
- 2,000,000
- TPM utilization
- 50.0%
- RPM utilization
- 5.0%
Providers cap tokens-per-minute (TPM) and requests-per-minute (RPM). Enter your traffic to see your utilization and whether you'll hit a limit.
How this is calculated
We compute your TPM usage (tokens per request × requests per minute) and RPM usage (requests per minute) as a percentage of your provider's limits, and flag whichever you'll hit first. Large prompts hit TPM first; many small calls hit RPM first.
Is this a good result? What to do next
The binding limit is whichever you use a higher percentage of. If you're near either, you'll see HTTP 429s under load — add backoff, batch, or request a quota increase before launch, not after.
Typical planning ranges
- TPM usage
- tokens/request × RPM
- Large prompts
- hit TPM first
- Many small calls
- hit RPM first
Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.
How to improve this number
- Batch requests to reduce RPM pressure.
- Trim tokens per request to ease TPM.
- Request a quota increase or spread load across keys/regions.
Common mistakes
- Planning to 100% of a limit with no burst headroom.
- Watching only one of TPM/RPM.
When to use a different approach
To size instances/concurrency, use the concurrency capacity planner. For batch discounts on offline jobs, use the batch API savings calculator.
Worked example (defaults)
With the default inputs above, here is the result:
- Tokens / minute used
- 1,000,000
- TPM limit
- 2,000,000
- TPM utilization
- 50.0%
- RPM utilization
- 5.0%
Frequently asked questions
What if I exceed a limit?+
Requests get throttled (HTTP 429). Add exponential backoff, batch requests, or request a quota increase.
Which limit binds first?+
Whichever you use a higher percentage of. Large prompts hit TPM first; many small calls hit RPM first.