AI Paper Atlas

Research map / Benchmarks / MLVU

MLVU leaderboard

62 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on MLVU. Best accuracy: 81.2 (VideoARM, CVPR 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

MLVU · accuracy

62 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
181.2VideoARM
o3+GPT-4.1
VideoARM: Agentic Reasoning over Hierarchical Memory for Long-Form Video Understanding
CVPR 2026 · Table 1
Video-language understanding
281Symphony
DeepSeek R1 + Seed 1.6 VL
Symphony: A Cognitively-Inspired Multi-Agent System for Long-Video Understanding
CVPR 2026 · Table 2
Video-language understanding
380SelectStream
Qwen3-VL-8B
What Should a Streaming Video Model Remember?
NeurIPS 2026 · Table 2
Video-language understanding
477.6ReMem
Qwen3-VL-8B
Reasoning with Memory: A Temporal Granularity-Adaptive Framework for Training-Free Long Video Understanding
ECCV 2026 · Table 1
Video-language understanding
576.2CoTR (Ours)
Qwen3-VL-4B
Towards Temporal Compositional Reasoning in Long-Form Sports Videos
ECCV 2026 · Table 3
Video-language understanding
675.6Tempo (4K budget)
Qwen3-VL-2B + Qwen3-LM-4B
Small Vision-Language Models are Smart Compressors for Long Video Understanding
ECCV 2026 · Table 1
Video-language understanding
775.5LLaVA-Video-7B + G2F-RAG
LLaVA-Video-7B
Graph-to-Frame RAG: Visual-Space Knowledge Fusion for Training-Free and Auditable Video Reasoning
CVPR 2026 · Table 2
Video-language understanding
875.1DeltaVid-Qwen3-VL-8B
Qwen3-VL-8B
DELTAVID: Enhancing Fine-Grained Spatiotemporal Perception with Cross-Video Differences
NeurIPS 2026 · Table 2
Video-language understanding
974.8STVFocus
InternVL3-8B
STVFocus: Query-guided Spatio-Temporal Visual Focusing for Video LLMs
ECCV 2026 · Table 1
Video-language understanding
1074.2Qwen2.5-VL-72B + ReAgent-V
Qwen2.5-VL-72B
ReAgent-V: A Reward-Driven Multi-Agent Framework for Video Understanding
NeurIPS 2025 · Table 1
Video-language understanding
52 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.