VideoZeroBench - Level-1 (Temporal and Spatial 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-1 accuracy, with the annotated time intervals and key spatial regions given in text alongside the video, in percent; higher is better.
Metric: Level-1 accuracy (%), answering with the annotated temporal intervals and key spatial regions 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 | 28.4 | #486 |
| 2 | Seed 2.0 Pro | 25.4 | #226 |
| 3 | Gemini 2.5 Pro | 25.2 | #280 |
| 4 | Qwen 3.5 397B A17B | 25 | #270 |
| 5 | Qwen 3 VL 8B | 24.8 | #927 |
| 6 | Gemini 3 Pro | 23.6 | #146 |
| 7 | Qwen 3 VL 4B | 22.2 | #1163 |
| 8 | InternVL3.5-8B | 19.4 | #1618 |
| 9 | Open-o3-Video-4B | 17.6 | |
| 10 | GPT-5.2 | 17.4 | #190 |
| 11 | VideoChat-R1.5-7B | 17.4 | #1260 |
| 12 | Video-R1 | 17.2 | #1384 |
| 13 | Gemini 2.5 Flash | 17 | #466 |
| 14 | Qwen 2.5 VL 7B | 16.2 | #1878 |
| 15 | Video-o3-7B | 15 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=videozerobench-level-1-temporal-and-spatial-evidence-given · How It Works · Data refreshed daily, snapshot 2026-11-22.