RESCUE-BENCH - Viewpoint Prediction: leaderboard
Metric: Judge Score (1-5). Source: arxiv.org. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4 Flash | 4.08 |
| 2 | Qwen 3.5 Plus | 3.94 |
| 3 | DeepSeek V4 Pro | 3.9 |
| 4 | DeepSeek V3.2 | 3.86 |
| 5 | Kimi K2.5 | 3.64 |
| 6 | DeepSeek R1 | 3.35 |
| 7 | Qwen 3 Max | 3.35 |
| 8 | MiniMax-M2.5 | 3.25 |
| 9 | GPT-4o | 3.24 |
| 10 | Qwen Plus | 3.2 |
Interactive version: theaggregate.ai/benchmark?slug=rescue-bench-viewpoint-prediction · How It Works · Data refreshed daily, snapshot 2026-09-21.