OmniCompass-AV - Question Answering - Consistency Score: leaderboard
OmniCompass-AV - Question Answering - Consistency Score. Matched text-only/text-plus-image target specifications; consistency requires both target forms correct. 240 audiovisual videos, 720 task instances. Source-specific frame budgets retained in primary table; no fusion with single-condition accuracy. Metric: Consistency score (%). Primary published table; source configurations remain distinct.
Metric: Consistency score (%). Source: arxiv.org. Saturation forecast: Around January 2027. 17 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.8 Flash | 78.2 | #21 |
| 2 | Qwen3.8 Omni Flash | 73.2 | #29 |
| 3 | Qwen 3.5 Omni Plus | 68.4 | #82 |
| 4 | Gemini 3.1 Pro (Preview) | 64.9 | #56 |
| 5 | Qwen3.5 Omni Flash | 58 | #173 |
| 6 | MiMo-V2.5 | 47.3 | #192 |
| 7 | Qwen2.5-Omni-7B | 43.8 | #608 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=omnicompass-av-question-answering-consistency-score · How It Works · Data refreshed daily, snapshot 2026-10-16.