Agentic-OPD Study - DeepSearch - HotpotQA: leaderboard
Agentic-OPD v1 Table1, unmodified base checkpoints only. DeepSearch exact-match percent, Wikipedia searchmax4turns, source-defined benchmarktestsets. RLtrained derivatives and unboundGPT5.6/DeepSeekV4.1 tiers omitted; no cross-generation inference.
Metric: Score (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 7B Instruct | 28.3 | #2583 |
| 2 | Qwen 3 4B 2507 Instruct | 24.1 | #2174 |
| 3 | Qwen 2.5 3B Instruct | 21.2 | #5206 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=agentic-opd-study-deepsearch-hotpotqa · How It Works · Data refreshed daily, snapshot 2026-11-12.