loading
loading
A hard cap on how many requests or tokens you can send to an API in a given window.
Providers enforce limits like requests-per-minute or tokens-per-minute to protect their infrastructure and your account, and crossing them returns a 429 error instead of a result. It matters because parallel agents and retry loops hit these ceilings fast, so production code needs backoff and queuing to stay under them. Example: firing 100 concurrent calls trips a 50-requests-per-minute limit, so you add exponential backoff to retry the rejected ones a few seconds later.
Plainly
Think of Rate Limit as the checklist that keeps the app open for real people. A hard cap on how many requests or tokens you can send to an API in a given window.
In practice
Use it when a change has to survive deploys, users, incidents, analytics, or billing reality. In practice, define the owner, input, output, and failure mode before you rely on it.