What AI says about GTA

GTA is a benchmark to evaluate the tool-use capability of LLM-based agents in real-world scenarios. It features three main aspects: real user queries (229 human-written prompts with implicit tool-use), real deployed tools across perception, operation, logic, and creativity categories, and real multimodal inputs (authentic images, tables, code, etc.). The project provides a dataset, paper, project page, Hugging Face, and a new leaderboard to compare agent performance.