What is 'API throttling vs rate limiting' and when would you use each for a public API?

• A.Rate limiting queues excess requests in a buffer (a bounded FIFO list, size 100) instead of rejecting them. Throttling rejects requests with a 429 once the window ends, rather than slowing them down; rate limiting provides graceful degradation, throttling protects backend resources by rejecting excess. Public APIs often use throttling for abuse prevention and rate limiting for fair resource distribution among legitimate users, with the limits stored in the gateway configuration
• B.Rate limiting restricts the number of requests in a time window (100 requests/minute - exceeding returns 429). Throttling degrades service quality for excess requests rather than rejecting them (slow down responses, prioritize lower-load clients). Rate limiting protects backend resources by rejecting excess; throttling provides graceful degradation. Public APIs often use rate limiting for abuse prevention and throttling for fair resource distribution among legitimate users