Can LLM Agents Respond to Disasters? Benchmark: leaderboard
Operational disaster response goes beyond damage assessment, requiring responders to integrate multi-sensor signals, reason over road networks, populations and key facilities, plan evacuations, and produce actionable reports. However, prior work largely isolates remote-sensing perception or evaluates generic tool use, leaving the end-to-end workflows of emergency operations underexplored. In this
Metric: AVG (self-reported). Source: benchmarklist.com. Saturation forecast: Around June 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 53.74 | #99 |
| 2 | Qwen 3.5 397B A17B | 53.45 | #146 |
| 3 | MiMo-V2-Pro | 52.89 | #168 |
| 4 | Grok 4.1 Fast | 52.1 | #217 |
| 5 | Claude Sonnet 4.6 | 52.01 | #90 |
| 6 | Gemma 4 31B | 51.17 | #200 |
| 7 | MiniMax-M2.7 | 48.35 | #258 |
| 8 | DeepSeek V3.2 | 48.23 | #205 |
| 9 | GPT-5.4 | 47.63 | #79 |
| 10 | Step 3.5 Flash | 46.68 | #270 |
| 11 | GPT-5.4 Nano | 38.14 | #328 |
| 12 | GPT-OSS-120B | 35.11 | #341 |
| 13 | Qwen 3.5 35B A3B | 24.01 | #256 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=can-llm-agents-respond-to-disasters-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.