A new AI Agent Leaderboard has been launched by Rungalileo, evaluating the performance of 17 large language models (LLMs) across 14 benchmarks. The leaderboard assesses models on their ability to utilize tools in complex scenarios, including single-turn and multi-turn interactions, as well as error handling. Leading models in this evaluation include Google's DeepMind Gemini-2.0-flash and OpenAI's GPT-4o. The leaderboard aims to provide insights into how AI agents perform in real-world business situations, with a stunning interface created using Gradio 5. The launch has garnered attention, with coverage from ZDNet highlighting the significance of this evaluation for understanding AI capabilities.