Rate limit requests from clients to protect the health of the service and maintain high service quality for other clients. You can use a token bucket algorithm to quantify request limits.
Return the remaining number of request tokens with each request in the
RateLimit-Remaining response header.