LongShOTBench: leaderboard
MBZUAI LongShOTBench (arXiv:2512.16978): long-form omni-modal video understanding benchmark combining visual, speech, and ambient-audio cues across open-ended single- and multi-turn questions graded with weighted criterion-level rubrics (0-100%). Overall score across standalone models (multi-model agentic frameworks excluded); higher is better.
Metric: Weighted Rubric Score (%). Source: longshot.cvmbzuai.com. Saturation forecast: Around 2028. 102 models tracked. Updated 21 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Qwen 3 Omni 30B A3B (Thinking) | 64.2 | ||
| 2 | Gemini 3.1 Pro (Preview) | 56.97 | #99 | |
| 3 | Voxtral-Mini-3B-2507 | 49.61 | #3682 | |
| 4 | Voxtral-Small-24B-2507 | 48.56 | #1246 | |
| 5 | Qwen 3 VL 235B A22B (Thinking) | 47.59 | #459 | |
| 6 | Qwen 3 VL 32B (Thinking) | 46.72 | #573 | |
| 7 | Vero-Qwen3I-8B | 46.18 | ||
| 8 | Kimi K2.6 | 45.07 | #192 | |
| 9 | Qwen 3 VL 30B A3B (Thinking) | 43.71 | #719 | |
| 10 | Qwen 3 VL 235B A22B Instruct | 42.58 | #532 | |
| 11 | Qwen 3 VL 8B (Thinking) | 42.18 | #910 | |
| 12 | Gemma 3 27B (IT) | 41.97 | #1324 | |
| 13 | Intern-S1-mini | 41.45 | #1178 | |
| 14 | Nemotron 3 Nano Omni 30B A3B | Reasoning | 41.22 | #1412 |
| 15 | Qwen 3 VL 32B Instruct | 40.59 | #606 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=longshotbench · How It Works · Data refreshed daily, snapshot 2026-11-21.