Amazon API Gateway is a managed service that scales for you automatically. You still need to be aware of the service quotas and request increase if you plan to go above your current ones.

It is important to keep in mind the entire application stack while you plan your scaling. It is no use having tens of thousands requests per second quota on API Gateway, if your backend integration cannot handle such load. Monitor your resource usage and quotas across the entire stack (e.g. AWS Lambda concurrency, Amazon DynamoDB provisioned capacity, etc.). We will talk about API Gateway monitoring in the Monitor lesson.

Note: Automatic quota management is available for AWS Service Quotas. This feature can adjust values of AWS services’ quotas automatically and safely based on customer’s usage

As your service scales globally, consider edge-optimized API Gateway endpoints or self-managed Amazon CloudFront distribution to improve client connection time. Also consider caching responses to the common data requests that do not change that often.

In addition, look at cell-based architectures and multi-account approaches for easier blast radius control and quota management. Use load shedding to avoid overload, including API Gateway request validation.