SciHazard (Deep Research Agents): leaderboard
SciHazard (Shanghai AI Laboratory, 2026) measures scientific misuse risk with 2,400 hazardous questions grounded in regulated substances and documented misuse scenarios across 12 disciplines, plus 600 benign questions that reuse the same trigger words. The DeHarm-Score multiplies each question hazard level by whether the answer is refused and by its judged executability and net-new risk, on a 0 to 5 scale where lower is safer; this board covers deep-research agents and agent frameworks with web retrieval.
Metric: DeHarm-Score (0-5, lower is safer; the mean over 2,400 hazardous questions in 12 scientific disciplines of hazard level 1 to 5 x non-refusal x (executability + net-new risk, each 1 to 4)/8, executability from importance-weighted checklists and net-new risk from retrieval-checked claims, judged by Qwen3.5-35B-A3B; a refused answer scores 0; deep-research agents and agent frameworks with web retrieval at provider defaults). Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Sonar Pro | 1.2 | #722 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=scihazard-deep-research-agents · How It Works · Data refreshed daily, snapshot 2026-11-13.