VideoZeroBench - Level-2 (Temporal Evidence Given): leaderboard
Metric: Level-2 accuracy (%), answering with only the annotated temporal intervals given in text, on VideoZeroBench's 500 manually annotated open-ended questions (exact short answers) over 138 long videos in 13 domains; Gemini models get the raw video through the API, the others 1 fps frames up to the printed cap, temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 235B A22B | 21.4 |
| 2 | Gemini 2.5 Pro | 21 |
| 3 | Gemini 3 Pro | 20.6 |
| 4 | Qwen 3.5 397B A17B | 20 |
| 5 | Seed 2.0 Pro | 18 |
| 6 | GPT-5.2 | 14.6 |
| 7 | Qwen 2.5 VL 7B | 12.4 |
| 8 | Gemini 2.5 Flash | 11.2 |
Interactive version: theaggregate.ai/benchmark?slug=videozerobench-level-2-temporal-evidence-given · How It Works · Data refreshed daily, snapshot 2026-10-07.