AIPricingLabGlossary
Glossary

Rate limit

A short-window cap on how often a request can be made - typically per-second or per-minute. Distinct from a quota, which is a long-window cap.

Last updated: 2026-05-10

Definition

Rate limits protect infrastructure from spikes (60 req/sec/user). Quotas enforce business rules (100 image renders / month / user). The two are complementary: rate limits live at your edge (Cloudflare, Vercel) and care about IP / second; quotas live in your business logic and care about user / month. AIPricingLab is built for quotas, not rate limits - you typically still want a Redis or edge-based rate limiter for sub-second protection.

Example

A free-tier user has rate limit 30 req/min (anti-abuse) AND quota 20 renders/month (cost cap). Both apply.

Related terms