VideoOdyssey-AV - 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 audio-visual subset questions about inferring emotions and their changes; higher is better.

Metric: Multiple-choice accuracy (%) on the 62 emotional reasoning questions (infer the emotional state, underlying causes, and evolutionary trajectories of specific entities or the overall atmosphere, based on both visual and audio cues) of VideoOdyssey-AV (human-written questions on 100 videos of 60 to 255 minutes that need both visual and audio cues; 1,062 in all), omni-modal models given video and audio, 64 frames (VITA-1.5-7B 16); higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 12 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Flash51.6#175
2Qwen 3.5 Omni Plus48.4#254
3Gemini 2.5 Pro46.8#276
4Gemini 3.1 Pro (Preview)46.8#99
5Ola-7B35.5#1557
6VideoLLaMA2-7B33.9#2385
7OneLLM30.7#2372
8Qwen 3 Omni 30B A3B29#847
9Unified-IO-2 XL29
10VITA-1.5-7B29#1853
11Unified-IO-2 L27.4
12Unified-IO-2 XXL22.6

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

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