VideoOdyssey-V - Intentional 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 motives; higher is better.

Metric: Multiple-choice accuracy (%) on the 66 intentional reasoning questions (infer the underlying purposes or motivations behind a specific character’s actions) 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 February 2027. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3.1 Pro (Preview)78.8#99
2Gemini 2.5 Pro62.1#276
3Kimi K2.562.1#266
4GPT-5.260.6#185
5Seed 2.0 Pro57.6#222
6Qwen 3.5 27B50#375
7Qwen 3 VL 235B A22B48.5#485
8Qwen 3 VL 32B40.9#622
9Video-KTR-7B37.9
10InternVL3.5-38B36.4#1021
11LLaVA-OneVision-1.5-8B36.4#1864
12Video-R134.8#1379
13LLaVA-NeXT-Video-DPO (34B)33.3#2489
14VideoLLaMA3-7B31.8#2050
15Phi-4 Multimodal Instruct28.8#4154

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-intentional-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.