A mechanism that restricts the number of requests a user, IP address, or API key can make within a specified time period. Protects against abuse, brute force attacks, and uncontrolled costs.
Why it matters
Protects the system from request floods: intentional attacks, runaway scripts, and bots. Establishes fair boundaries so that one client cannot exhaust resources for everyone else.
What's missing without it
Without limits, a single script could bring down an API, exhaust expensive AI call quotas, or brute-force test a million passwords on a login form.
When it is used
On every public endpoint; especially for login, forms, and costly operations. Used in two areas: securing public routes and protecting forms and login processes.
How we use it
Public API allows 60 requests per minute per integration key. Typical values: five login attempts per minute from one IP address, 100 API requests per minute per key.
Numbers worth knowing
Rate limiting in data
29 300 000
Liczba internautów w Polsce w sierpniu 2026 według badania Mediapanel (lipiec 2026? - w materiale jest lipiec, ale mowa o sierpniu, dlatego podaję jak w tekście: 'In July, the number of Internet users amounted to 29.3 million')
Mediapanel09/2026Poland