EmoMM - Align: leaderboard
The Align subset of EmoMM (Sun et al., 2026): test clips in which the text, audio and video sentiment all agree with the overall label. Five-class sentiment accuracy in percent; higher is better.
Metric: Accuracy (%; EmoMM-Align test subset, every modality's polarity agrees with the ground truth; five-class sentiment). Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 59.85 | #135 |
| 2 | Qwen2.5-Omni-7B | 52.09 | #1797 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=emomm-align · How It Works · Data refreshed daily, snapshot 2026-11-12.