Security and Limits for API Designs
API security and limits cover who may call an API and how much they may call it: authentication with API keys, OAuth 2.0 or JWTs, authorisation per resource, and rate limiting with quotas. The design questions are which scheme fits which client and how limits are communicated back to callers.
Stage: Practical Design · Part of the API Design roadmap · Last reviewed 19 November 2026
What to learn in Security and Limits
- Choose between API keys, OAuth 2.0 and mutual TLS for a given client type
- Apply authorisation checks per resource so users cannot read other users' objects
- Design rate limits and return 429 with a Retry-After header
- Expose quota information to clients through response headers
- Explain why HTTPS is required and how tokens should be scoped
Security and Limits interview questions
What does HTTP 429 mean and how should a client react?
429 Too Many Requests means the caller exceeded a rate limit. The response should include a Retry-After header or similar reset information. A well-behaved client stops sending, waits for that time, then retries with exponential backoff and some random jitter.
What is the difference between authentication and authorisation in an API?
Authentication proves who the caller is, for example by validating a token. Authorisation decides what that identity may do, such as read one order but not another customer's order. APIs often get the first right and fail the second, which causes broken object-level access flaws.
Where Security and Limits fits in the API Design roadmap
Security and Limits belongs to the Practical Design stage (intermediate level), which takes about 15 hours. Everyday decisions.
Other units in this stage
Back to the full API Design roadmap