DisasterBench - Chain-of-Thought: leaderboard

Chain-of-thought planning, with intermediate reasoning before the plan, on DisasterBench (2026), which tests whether an LLM can compose executable disaster-response workflows from 26 specialized tools under typed interface constraints. Exact-match accuracy over the planning tasks (every step must match in tool, parameter bindings and dependencies), percent; higher is better.

Metric: Exact-match accuracy (%) over the 233 DisasterBench planning tasks (35 single-tool, 166 chain and 32 branching workflows over 26 disaster-response tools): a plan counts only when every step matches the ground truth in tool identity, parameter bindings and inter-step dependencies; planning method: intermediate reasoning before the final plan (temperature 0); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3.1 Pro (Preview)72.53#99
2Qwen 3 MaxThinking66.52#415
3Gemma 4 31B63.09#382
4DeepSeek R162.66#482
5GPT-5.462.23#135
6Qwen 3.5 27B62.23#375
7Qwen 3.5 9B61.37#828
8Llama 3.3 70B58.8#1514
9DeepSeek V3.236.05#419
10Ministral 3 14B33.48#1870
11Ministral 8B (DisasterBench checkpoint unspecified)31.33
12Llama 3.1 8B26.18#5196
13Ministral 3B (DisasterBench checkpoint unspecified)21.03

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=disasterbench-chain-of-thought · How It Works · Data refreshed daily, snapshot 2026-11-21.