VideoOdyssey-V - Emotional Reasoning: leaderboard

VideoOdyssey (2026) tests ultra-long video understanding with human-written multiple-choice questions on videos averaging 109 minutes, each annotated with how long a viewer must watch continuously to answer. This board is accuracy in percent on the visual subset questions about inferring emotions and their changes; higher is better.

Metric: Multiple-choice accuracy (%) on the 71 emotional reasoning questions (infer the emotional state, underlying causes, and evolutionary trajectories of specific entities or the overall atmosphere) of VideoOdyssey-V (human-written questions on 100 videos of 60 to 255 minutes that need visual cues and long continuous viewing; 1,618 in all), 128 frames for proprietary models and each open model's maximum frame setting; higher is better. Source: arxiv.org. Saturation forecast: Around October 2028. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Kimi K2.552.1#266
2Gemini 2.5 Pro49.3#276
3Gemini 3.1 Pro (Preview)45.1#99
4GPT-5.243.7#185
5Seed 2.0 Pro43.7#222
6Qwen 3.5 27B42.3#375
7InternVL3.5-38B39.4#1021
8Video-R139.4#1379
9Video-KTR-7B36.6
10VideoLLaMA3-7B35.2#2050
11Qwen 3 VL 235B A22B33.8#485
12Qwen 3 VL 32B33.8#622
13LLaVA-OneVision-1.5-8B33.8#1864
14Phi-4 Multimodal Instruct31#4154
15LLaVA-NeXT-Video-DPO (34B)31#2489

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=videoodyssey-v-emotional-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.