LiveAgentBench: Comprehensive Benchmarking of Agentic Systems Across 104 Real-World Challenges
This work addresses the limitation of existing benchmarks for general AI agents, providing a more accurate representation of real-world user tasks for researchers and developers in the AI community.
The authors tackled the problem of benchmarking general AI agents with a comprehensive benchmark called LiveAgentBench, which consists of 104 real-world scenarios and reveals the practical performance of various models and frameworks. The benchmark includes 374 tasks and enables continuous updates with fresh queries from real-world interactions.
As large language models grow more capable, general AI agents have become increasingly prevalent in practical applications. However, existing benchmarks face significant limitations, failing to represent real-world user tasks accurately. To address this gap, we present LiveAgentBench, a comprehensive benchmark with 104 scenarios that reflect real user requirements. It is constructed from publicly sourced questions on social media and real-world products. Central to our approach is the Social Perception-Driven Data Generation (SPDG) method, a novel process we developed to ensure each question's real-world relevance, task complexity, and result verifiability. We evaluate various models, frameworks, and commercial products using LiveAgentBench, revealing their practical performance and identifying areas for improvement. This release includes 374 tasks, with 125 for validation and 249 for testing. The SPDG process enables continuous updates with fresh queries from real-world interactions.