VANTAGE-Bench: leaderboard
VANTAGE-Bench evaluating vision-language models on spatial, temporal, and semantic video understanding (overall %).
Metric: Overall (%). Source: huggingface.co. Saturation forecast: Around March 2027. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 75.26 | #6 |
| 2 | Gemini 3.6 Flash | 69.52 | #51 |
| 3 | GPT-5.6 Sol | 68.04 | #19 |
| 4 | Gemini 3.1 Pro (Preview) | 62.66 | #58 |
| 5 | Qwen 3 VL 32B Instruct | 55.38 | #283 |
| 6 | Gemini 3.5 Flash Lite | 53.25 | #156 |
| 7 | Gemini 3.1 Flash Lite | 52.61 | #191 |
| 8 | Qwen 3 VL 8B Instruct | 50.07 | #420 |
| 9 | Qwen 3.5 27B | 49.25 | #187 |
| 10 | Qwen 3.5 9B | 48.3 | #379 |
| 11 | Gemma 4 E2B (IT) | 21.28 | #781 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=vantage-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.