API Rate Limit Calculator
Turn an API rate limit into sustainable throughput per hour and day, and the time to complete a batch of requests.
How it works
Sustainable rate = limit ÷ window · Windows needed = ⌈total ÷ limit⌉ · Time ≈ (windows − 1) × window
A rate limit of N requests per window sets a hard ceiling on sustained throughput of N ÷ window requests per second. To push a large batch through, you split it into full windows: the number of windows is the total divided by the limit, rounded up. The elapsed time is roughly one window less than that, because the first batch can go out immediately and you only wait between windows.
Worked example
A limit of 100 requests per 60-second window is 100 ÷ 60 ≈ 1.67 req/s, or about 6,000 per hour. To send 10,000 requests you need ⌈10,000 ÷ 100⌉ = 100 windows, so you wait 99 windows × 60 s ≈ 5,940 s, roughly 99 minutes.
Frequently asked questions
Should I run right at the limit?
No. Leave headroom — clock skew, retries and bursts can push you over and trigger 429 responses or bans. Targeting 80–90% of the limit and honoring any Retry-After header is safer.
What is the difference between a hard limit and a burst?
Many APIs use a token-bucket scheme that allows short bursts above the average rate as long as the long-run average stays under the limit. This calculator models the sustained average; a burst allowance lets you briefly exceed the per-second figure.
How do concurrent requests affect this?
Concurrency changes latency, not the ceiling. If the limit is 100 per minute, running 10 requests in parallel still consumes 10 of your 100 allowed calls — you finish a batch faster but hit the same window cap.
Related calculators
This calculator is for educational purposes. Double-check important results before acting on them.