Benchmark
OmniCapBench · Structural · Schema Gate Compliance
Scores are percentages. Higher is better.
Benchmark sourceScores from BenchBoard docs.
- Models and systems measured
- 9
- Scores recorded
- 9
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 91.3%
- Mean
- 95.6%
- Median
- 7.6%
- Std dev
- 87.7%
- 10th percentile
- 96.7%
- 90th percentile
- 72.1%
- Min
Leaderboard
| # | Model or system | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro PreviewOmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 97.4% |
| 2 | MiMo V2.5OmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 96.6% |
| 3 | Qwen3.5 Omni PlusOmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 96.1% |
| 4 | Qwen3.5 Omni FlashOmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 95.6% |
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| Gemini 2.5 ProOmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 96.7% |
| MiniCPM-o 2.6 (OmniCapBench label)OmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 72.1% |
| Qwen3 Omni Captioner (OmniCapBench label)OmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 90.8% |
| Qwen3 Omni Instruct (OmniCapBench label)OmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 89% |
| Seed2.0 (OmniCapBench label)OmniCapBench benchmark-owner run · 786 audio-video clips (12.8 hours) · structured MTSS captions · macro average across videos · rule-based structural track with LLM-assisted reference matching · candidate reasoning effort, exact API snapshot and judge model not specified on public leaderboard· Reported by OmniCapBench authors / Tencent Hunyuan· via BenchBoard docs | 87.7% |