VibeSearchBench: Benchmarking Long-horizon Proactive Search in the Wild
This benchmark exposes a critical evaluation-experience gap for LLM-based search agents, highlighting the need for advances in proactive intent elicitation and multi-turn reasoning.
LLM-based agents perform well on existing search benchmarks but fail in real-world use due to over-specified queries and single-turn interactions. The authors introduce VibeSearchBench, a benchmark with 200 bilingual tasks, and find that even the best model achieves only 30.30 F1, revealing a significant gap in long-horizon proactive search capabilities.
LLM-based agents score well on search benchmarks, yet real users consistently find results unsatisfying, revealing a persistent evaluation-experience gap. We attribute this gap to existing benchmarks' reliance on over-specified queries, single-turn interactions, and fixed-schema evaluation, none of which reflect real search behavior where users and agents collaboratively refine vague intent through multi-turn dialogue. We term this paradigm VibeSearch and introduce VibeSearchBench, a benchmark comprising 200 manually curated bilingual (Chinese and English) tasks across 20 domains, split into VibeSearch-Pro (professional) and VibeSearch-Daily (daily-life) subsets. Each task pairs a user persona with a schema-free ground-truth knowledge graph, and is evaluated through a progressive-disclosure user simulator and a graph-matching evaluation framework. We benchmark seven frontier models under both the ReAct framework and the OpenClaw agent harness. Results show that all models remain substantially inadequate for VibeSearch (best F1: 30.30), highlighting the need for fundamental advances in long-context reasoning, proactive intent elicitation, and structured knowledge construction.