VEBench - Footage Selection: leaderboard
The footage selection task of VEBench (2026): choose which of four candidate clips best continues a reference video as A-roll and B-roll editing would, with the candidates concatenated and marked on screen and subtitles provided. Accuracy in percent; higher is better.
Metric: Footage Selection accuracy (%) on Operation Simulation: four-option choice of the candidate clip that best fits a reference video (A/B-roll pairing), candidates concatenated into one sequence with on-screen text markers, subtitles provided; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 9 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 44.44 | #276 |
| 2 | GPT-4o | 41.89 | #751 |
| 3 | Qwen 3 VL 32B Instruct | 39.72 | #606 |
| 4 | Qwen 3 VL 8B Instruct | 36.14 | #1029 |
| 5 | Qwen 3 VL 4B Instruct | 31.8 | #1332 |
| 6 | InternVL3-14B | 31.16 | #1214 |
| 7 | InternVL3-8B | 30.78 | #1782 |
| 8 | Qwen 2.5 VL 7B Instruct | 30.01 | #1902 |
| 9 | Qwen 3 VL 30B A3B Instruct | 29.37 | #895 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=vebench-footage-selection · How It Works · Data refreshed daily, snapshot 2026-11-21.