Users interact with applications at the UI level while developers and testers are responsible for ensuring the reliability of underlying APIs. Without testing APIs, testers would be limited to testing at the UI level, which requires waiting until the complete application stack is built before beginning.
Performing API testing at the API level allows testers and developers to design test cases that interact directly with underlying APIs. This makes it possible to test business logic in a stable manner at an easy-to-automate layer. UI testing limits testers to validating a specific user experience. API testing allows testers to protect applications against the unknown.