EmoTrans: leaderboard
EmoTrans (2026) tests whether multimodal LLMs follow emotion transitions in real-world videos: detecting a change, naming the emotions before and after, and predicting the next emotion, over 1,000 annotated clips. Average accuracy (%) of the three classification tasks; higher is better.
Metric: Average accuracy (%) over Emotion Change Detection, Emotion State Identification and Next Emotion Prediction on EmoTrans, built from 1,000 manually annotated real-world video clips (OpenHumanVid and CH-SIMS v2) over 12 scenarios with one or more target persons; zero-shot, video with its text (and audio where the model accepts it); higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 15 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-5.4 | 49.3 | #135 | |
| 2 | Seed 2.0 Lite | 47.4 | #300 | |
| 3 | Qwen 3.5 122B A10B | 47 | #346 | |
| 4 | Gemini 3 Flash | Non-reasoning | 46.8 | #175 |
| 5 | Gemini 3 Pro | 46.7 | #145 | |
| 6 | Qwen 3.5 27B | 46.4 | #375 | |
| 7 | GLM-4.6V | 46.1 | #710 | |
| 8 | GLM-4.5V | 45.2 | #754 | |
| 9 | Qwen 3 VL 32B Instruct | 43.5 | #606 | |
| 10 | Qwen 2.5 VL 72B Instruct | 43.5 | #841 | |
| 11 | Gemini 2.5 Flash | Non-reasoning | 40.9 | #463 |
| 12 | Qwen 3 VL 8B | 40.9 | #917 | |
| 13 | InternVL3.5-14B | 37.8 | #1504 | |
| 14 | Qwen 3 Omni 30B A3B | 37.4 | #847 | |
| 15 | InternVL3.5-8B | 37.2 | #1604 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=emotrans · How It Works · Data refreshed daily, snapshot 2026-11-21.