LongShOTBench - Short Videos: leaderboard

LongShOTBench (arXiv:2512.16978) Short Videos duration slice: weighted criterion-level rubric score (0-100%) on short-duration omni-modal videos; higher is better.

Metric: Weighted Rubric Score (%). Source: longshot.cvmbzuai.com. Saturation forecast: Around 2029. 102 models tracked. Updated 21 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen 3 Omni 30B A3B (Thinking)58.05
2Gemini 3.1 Pro (Preview)48.53#99
3Qwen 3 VL 235B A22B Instruct43.47#532
4Vero-Qwen3I-8B42.57
5Qwen 3 VL 235B A22B (Thinking)42.56#459
6Voxtral-Mini-3B-250742.26#3682
7Voxtral-Small-24B-250742.13#1246
8Qwen 3 VL 32B (Thinking)41.73#573
9Nemotron 3 Nano Omni 30B A3BReasoning40.76#1412
10Qwen 3 VL 32B Instruct40.23#606
11Qwen 3 VL 8B (Thinking)40.08#910
12Kimi K2.639.97#192
13Gemma 3 27B (IT)38.26#1324
14Qwen 3 VL 30B A3B (Thinking)36.49#719
15Gemma 4 31B (IT)36.38#350

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=longshotbench-short-videos · How It Works · Data refreshed daily, snapshot 2026-11-21.