Paper Detail

Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding

Chaoyou Fu, Haozhi Yuan, Yuhao Dong, Yi-Fan Zhang, Yunhang Shen, Xiaoxing Hu, Xueying Li, Jinsen Su, Chengwu Long, Xiaoyao Xie, Yongkang Xie, Xiawu Zheng, Xue Yang, Haoyu Cao, Yunsheng Wu, Ziwei Liu, Xing Sun, Caifeng Shan, Ran He

huggingface Score 14.0

Published 2026-04-06 · First seen 2026-04-08

General AI

Abstract

With the rapid advancement of video understanding, existing benchmarks are becoming increasingly saturated, exposing a critical discrepancy between inflated leaderboard scores and real-world model capabilities. To address this widening gap, we introduce Video-MME-v2, a comprehensive benchmark designed to rigorously evaluate the robustness and faithfulness of video understanding. To systematically evaluate model capabilities, we design a progressive tri-level hierarchy that incrementally increases the complexity of video comprehension, ranging from multi-point visual information aggregation, to temporal dynamics modeling, and ultimately to complex multimodal reasoning. Besides, in contrast to conventional per-question accuracy, we propose a group-based non-linear evaluation strategy that enforces both consistency across related queries and coherence in multi-step reasoning. It penalizes fragmented or guess-based correctness and assigns credit only to answers supported by valid reasoning. To guarantee data quality, Video-MME-v2 is constructed through a rigorously controlled human annotation pipeline, involving 12 annotators and 50 independent reviewers. Backed by 3,300 human-hours and up to 5 rounds of quality assurance, Video-MME-v2 aims to serve as one of the most authoritative video benchmarks. Extensive experiments reveal a substantial gap between current best model Gemini-3-Pro and human experts, and uncover a clear hierarchical bottleneck where errors in visual information aggregation and temporal modeling propagate to limit high-level reasoning. We further find that thinking-based reasoning is highly dependent on textual cues, improving performance with subtitles but sometimes degrading it in purely visual settings. By exposing these limitations, Video-MME-v2 establishes a demanding new testbed for the development of next-generation video MLLMs.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@misc{fu2026video,
  title = {Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding},
  author = {Chaoyou Fu and Haozhi Yuan and Yuhao Dong and Yi-Fan Zhang and Yunhang Shen and Xiaoxing Hu and Xueying Li and Jinsen Su and Chengwu Long and Xiaoyao Xie and Yongkang Xie and Xiawu Zheng and Xue Yang and Haoyu Cao and Yunsheng Wu and Ziwei Liu and Xing Sun and Caifeng Shan and Ran He},
  year = {2026},
  abstract = {With the rapid advancement of video understanding, existing benchmarks are becoming increasingly saturated, exposing a critical discrepancy between inflated leaderboard scores and real-world model capabilities. To address this widening gap, we introduce Video-MME-v2, a comprehensive benchmark designed to rigorously evaluate the robustness and faithfulness of video understanding. To systematically evaluate model capabilities, we design a progressive tri-level hierarchy that incrementally increase},
  url = {https://huggingface.co/papers/2604.05015},
  keywords = {video understanding, video MLLMs, progressive tri-level hierarchy, group-based non-linear evaluation, human annotation pipeline, multimodal reasoning, visual information aggregation, temporal dynamics modeling, high-level reasoning, hierarchical bottleneck, thinking-based reasoning, textual cues, subtitles, code available, huggingface daily},
  eprint = {2604.05015},
  archiveprefix = {arXiv},
}

Metadata

{}