VideoZeroBench - Level-2 (Temporal Evidence Given): leaderboard
VideoZeroBench (2026) asks 500 manually annotated, deliberately hard open-ended questions with exact short answers about 138 long videos in 13 domains, each with annotated temporal intervals and bounding boxes as supporting evidence. Level-2 accuracy, with only the annotated time intervals given in text alongside the video, in percent; higher is better.
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 2029. 17 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 235B A22B | 21.4 | #485 |
| 2 | Gemini 2.5 Pro | 21 | #276 |
| 3 | Gemini 3 Pro | 20.6 | #145 |
| 4 | Qwen 3.5 397B A17B | 20 | #264 |
| 5 | Seed 2.0 Pro | 18 | #222 |
| 6 | InternVL3.5-8B | 18 | #1604 |
| 7 | Qwen 3 VL 8B | 17.8 | #917 |
| 8 | Qwen 3 VL 4B | 16.6 | #1151 |
| 9 | GPT-5.2 | 14.6 | #185 |
| 10 | InternVL3.5-4B | 13.6 | #1930 |
| 11 | Video-R1 | 13.4 | #1379 |
| 12 | Video-o3-7B | 13.4 | |
| 13 | Open-o3-Video-4B | 12.8 | |
| 14 | Qwen 2.5 VL 7B | 12.4 | #1862 |
| 15 | VideoChat-R1.5-7B | 12.2 | #1254 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=videozerobench-level-2-temporal-evidence-given · How It Works · Data refreshed daily, snapshot 2026-11-21.