Gorilla Benchmark API Bench

Gorilla Benchmark API Bench evaluates Large Language Models' ability to accurately use APIs and tools through function calling tasks. Developed at UC Berkeley, this benchmark tests models' capability to understand API documentation, select appropriate functions, and generate correct function calls. Gorilla API Bench measures tool usage accuracy and helps assess LLMs' practical utility in real-world API integration scenarios.