SLVMBench - Mistake Detection: leaderboard
Metric: Accuracy (%) on the mistake detection questions of SLVMBench in the long-memory setting (tutorial, then 10 minutes to 2 hours of distractor videos, then the target video). Source: arxiv.org. Saturation forecast: Around December 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 60.89 |
| 2 | GPT-5.2 | 45.04 |
| 3 | Qwen2.5-Omni-7B | 43.13 |
| 4 | GPT-4o (2024-05-13) | 30.61 |
Interactive version: theaggregate.ai/benchmark?slug=slvmbench-mistake-detection · How It Works · Data refreshed daily, snapshot 2026-09-29.