Plan rate limits for an API

Prompt that plans request limits for an API, with a method, the response when a limit is exceeded and how to announce it. Paste it into an AI tool of your choice, which then plans the rate limits.

Topic: Apps & prototypes
Type: Prompt
Level: Intermediate
Time: about 8 min

Results depend on the tool, the model and what you enter.

Result preview

Example: written by our editors, not real AI output.

Limit per API key (partners use fixed keys, IP addresses change). Limits – Read: 600 requests per minute – Write: 120 requests per minute – Search: 60 requests per minute Method: token bucket, allows short bursts. Drawback: slightly more computation per request. When exceeded: 429, headers RateLimit-Remaining and Retry-After, message "Too many requests. Please try again in 20 seconds." Assumption: peaks of 400 per second are spread across at least ten partners.

Fill in

What you type stays in this browser window. We store and send nothing.

APIWhich API, for what?

CallersWho calls the API?

LoadWhat do you know about the load?

Your prompt

Plan rate limiting for the following API.
API and purpose: [API]
Who calls it: [Callers]
Known load figures: [Load]
Approach: 1. Decide what to limit by, for example per key, per account or per IP address, and justify the choice. 2. Suggest limits per group of endpoints, separately for read and write requests, and derive them from the load figures. 3. Choose a method, such as a fixed window, a sliding window or a token bucket, and name its drawback. 4. Describe the response when a limit is exceeded: status code 429, headers with remaining requests and wait time, a clear error message. 5. Explain how limits are monitored, announced in advance and raised for individual callers. Do not invent load figures; where figures are missing, state your assumptions explicitly.

Take particular care to derive every limit from load figures or explicit assumptions and give callers clear feedback.

Before you answer, check: Are all instructions met? If key information is missing or a placeholder in square brackets is still in the text, ask instead of guessing. Mark assumptions as assumptions.

Plan rate limiting for the following API.
API and purpose: [API]
Who calls it: [Callers]
Known load figures: [Load]
Approach: 1. Decide what to limit by, for example per key, per account or per IP address, and justify the choice. 2. Suggest limits per group of endpoints, separately for read and write requests, and derive them from the load figures. 3. Choose a method, such as a fixed window, a sliding window or a token bucket, and name its drawback. 4. Describe the response when a limit is exceeded: status code 429, headers with remaining requests and wait time, a clear error message. 5. Explain how limits are monitored, announced in advance and raised for individual callers. Do not invent load figures; where figures are missing, state your assumptions explicitly.

Take particular care to derive every limit from load figures or explicit assumptions and give callers clear feedback.

Before you answer, check: Are all instructions met? If key information is missing or a placeholder in square brackets is still in the text, ask instead of guessing. Mark assumptions as assumptions.

This is an instruction for an AI, not the result itself.

Where this prompt works

These AI tools are a good fit for this prompt:

How to use it

Fill in the fields or tap a suggestion. Leave out names, addresses and other personal data.
Copy the finished prompt into an AI tool of your choice.
First switch the limits on in observe-only mode, which logs breaches but blocks nothing, and adjust them afterwards.

Why does this work?

Going from "limit by what" to "how much" to "what happens then" avoids limits that sound sensible but hit the wrong callers. Headers and clear error messages let callers adapt instead of simply retrying.

Tips

Expensive endpoints such as search or export usually need their own, lower limits than simple reads.

You may use our prompts freely, including at work. Please do not republish them as your own collection. We give no warranty for AI results; check them yourself. Terms of use

Plan rate limits for an API

Result preview

Fill in

Your prompt

Where this prompt works

How to use it

Why does this work?

Tips

Also a good fit

Choose pagination for an API list New

Write an API reference people can use right away New

Plan a throttled data backfill New