RESCUE-BENCH - Viewpoint Prediction: leaderboard
RESCUE-BENCH viewpoint prediction: articulating each party's perspective, scored 1-5 by an LLM judge. Higher is better. Source: arxiv.org/abs/2609.09657.
Metric: Judge Score (1-5). Source: arxiv.org. Saturation forecast: Around February 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Flash | 4.08 | #155 |
| 2 | Qwen 3.5 Plus | 3.94 | #133 |
| 3 | DeepSeek V4 Pro | 3.9 | #112 |
| 4 | DeepSeek V3.2 | 3.86 | #210 |
| 5 | Kimi K2.5 | 3.64 | #152 |
| 6 | DeepSeek R1 | 3.35 | #262 |
| 7 | Qwen 3 Max | 3.35 | #213 |
| 8 | MiniMax-M2.5 | 3.25 | #316 |
| 9 | GPT-4o | 3.24 | #357 |
| 10 | Qwen Plus | 3.2 | #393 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=rescue-bench-viewpoint-prediction · How It Works · Data refreshed daily, snapshot 2026-11-05.