VideoOdyssey-V - Spatial 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 reasoning about locations and paths; higher is better.

Metric: Multiple-choice accuracy (%) on the 61 spatial reasoning questions (reason about the spatial locations of objects, including relative directions and the paths between them) 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 2030. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3.1 Pro (Preview)50.8#99
2Seed 2.0 Pro47.5#222
3GPT-5.245.9#185
4Gemini 2.5 Pro41#276
5Qwen 3.5 27B41#375
6Kimi K2.539.3#266
7Phi-4 Multimodal Instruct37.7#4154
8Qwen 3 VL 235B A22B36.1#485
9Qwen 3 VL 32B36.1#622
10Video-R136.1#1379
11Video-KTR-7B31.1
12InternVL3.5-38B29.5#1021
13Video-LLaVA-7B27.9#3879
14LLaVA-OneVision-1.5-8B26.2#1864
15LLaVA-NeXT-Video-DPO (34B)24.6#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-spatial-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.