Token Rate Calculator
Type an RPM limit, a TPM limit, and how many tokens one request uses. The page says how many of those requests fit in one minute.
The numbers are calculated in the browser. Nothing is uploaded. The numbers are stored in the URL.
Leave RPM or TPM blank when that limit does not apply. A limit of 0 is rejected.
No rate
Per minute
About this tool
Tokens per request are the input count plus the output count. The page adds those two numbers. It does not know which tokens a provider counts. If your provider counts only input tokens, put 0 in output. The TPM ceiling is the TPM limit divided by tokens per request, rounded down. The RPM ceiling is the RPM limit you typed.
Allowed requests per minute is the smaller of the ceilings you set. A blank box means that limit is not applied. Both boxes blank is an error. A typed 0 is an error, because a limit of 0 allows no requests. When the two ceilings are equal, both limits bind. When one request uses more tokens than the TPM limit, the TPM ceiling is 0 and the batch does not fit in one minute.
Batch tokens are the request count times tokens per request. Minutes are the request count divided by the allowed rate. When that divides evenly, the page shows the whole number. Otherwise it shows the exact ratio and the steady-rate division. That division is not a promise about a fixed window or a sliding window. This page does not look up a provider's RPM or TPM.
Read more
Free books, courses, and software
Get my programming books, course editions, and complete software.
Get the download library →