Free tool / build
How many users can you handle?
Enter your provider's limits and your traffic. See how many users you can serve before you hit 429 errors.
No sign-up / runs in your browser
Do the maths before the 429s.
Verdict
You would hit your limit
0.06× your 100 users at peak
Users you can serve
6.2
In your busiest minute. 12.3 at average traffic.
Limit you hit first
tokens
Allows 69 calls a minute
Calls per action
4.2
Including retries
What your 100 users need at peak
- Requests per minute
- 1,120of 500Over
- Tokens per minute
- 3,248,000of 200,000Over
How to get more room
- Ask for a higher tier, or spread load over more than one key or region.
- Cut input tokens: shorter prompts and summarised history raise the token limit’s reach.
- Queue and smooth bursts instead of firing every call at once, and retry with growing waits.
- Send easy steps to a smaller model. Limits are usually separate per model.
This is a planning estimate. Providers also limit per day, per model and per organisation, and some count your requested maximum reply length rather than the real one.
Providers set limits per model, per tier and sometimes per day or organisation, and some estimate your reply length from the maximum you request. The defaults here are example numbers, not any provider's real limits. Enter your own.
How it works
Three steps, no sign-up.
- 01
Enter your limits
Requests per minute and tokens per minute from your provider's dashboard.
- 02
Describe one user action
Calls per action, tokens per call and how often a failed call is tried again.
- 03
Read the verdict
See which limit you hit first, how many users fit, and what your target traffic needs.
Questions people ask
Where do I find my rate limits?
On your provider's dashboard, usually under limits, usage tiers or rate limits. They depend on your account tier and the model.
Which limit matters, requests or tokens?
Whichever is hit first. Agents with long prompts usually hit the token limit; apps with many tiny calls hit the request limit. The calculator tells you which.
Why include retries?
A call that fails with a 429 still counts against your limit when you try again. A 5% retry rate means 5% more traffic than users alone create.
What is the busiest-minute factor?
Real traffic is bursty. If your busiest minute is twice the average, you need limits for twice the average load.
How do I see my real usage against limits?
Tracify records every call and every 429, so you see how close you run to your limits and when you were throttled.
More free tools
- Agent log checker
Paste a log or trace. See where it failed, what was slow and what wasted tokens.
- Token counter
Paste text. See the token count and price on GPT, Claude, Gemini and Llama.
- Model price comparison
Every model's price in one table, plus what your app costs per month on each.
- Prompt cost checker
Paste a prompt. Get a shorter, cheaper version and what it costs on each model.
Now measure the real thing.
Send a first trace and see the true cost, speed and failures of every run.