SearchGymBench - Simple QA: leaderboard
Tool-based search over a synthetic world from SearchGymBench (Zhang, He, Zhu, Wu, Yu, Chu, Zhang, Tan and Jia, 2026): questions about fictional entities whose answers must be found with search and page-access tools in a generated Wikipedia-style corpus, so memorised knowledge cannot help. Accuracy in percent on the simple multi-hop questions (462 in the released test set), judged by Qwen2.5-72B-Instruct with up to 64 agent turns; higher is better.
Metric: Accuracy (%; multi-hop questions over a synthetic Wikipedia-style corpus answered with search and page-access tools, Qwen2.5-72B-Instruct judge; up to 64 turns). Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.2 | 69.4 | #403 |
| 2 | Kimi K2 | 52.6 | #435 |
| 3 | Qwen 3 4B | 23.5 | #2297 |
| 4 | Qwen 3 8B | 17.6 | #1838 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=searchgymbench-simple-qa · How It Works · Data refreshed daily, snapshot 2026-11-12.