VideoEval-Pro: leaderboard
Each model watches uniformly sampled frames of long videos (10 minutes and over, 38 minutes on average) and gives a short free-text answer to each of 1,289 questions; the score is the share a GPT-4o-0806 judge marks correct, at the model's best frame count; the rows are the paper's runs. The questions are Video-MME, MLVU, LVBench and LongVideoBench multiple-choice items rewritten without their options, so guessing among choices is not possible.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 16 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-4.1 | 40.8 | #505 | |
| 2 | GPT-4.1 Mini | 39.9 | #793 | |
| 3 | Gemini 2.5 Flash (Preview 04-17) | Non-reasoning | 36.3 | #455 |
| 4 | Qwen 2.5 VL 7B Instruct | 27.7 | #1893 | |
| 5 | InternVideo2.5-Chat-8B | 27.2 | #2031 | |
| 6 | VideoChat-Flash-Qwen2-7B (res448) | 27 | ||
| 7 | Qwen 2 VL 7B Instruct | 26.5 | #2594 | |
| 8 | InternVL3-8B | 24.7 | #1756 | |
| 9 | InternVL2.5-8B | 24.6 | #2401 | |
| 10 | LLaVA-Video-7B-Qwen2 | 24.2 | #1910 | |
| 11 | LongVU-7B | 22.1 | ||
| 12 | Video-XL | 18.6 | #2583 | |
| 13 | Phi-4 Multimodal Instruct | 16.5 | #4209 | |
| 14 | LongVA-7B | 16.5 | #3414 | |
| 15 | Mantis-8B-Idefics2 | 14.8 | #6896 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=videoeval-pro · How It Works · Data refreshed daily, snapshot 2026-11-17.