Share your thoughts, 1 month free Claude Pro on usSee more

Video Understanding on MLVU (Accuracy)

81.5Accuracy

RETOOL-VIDEO

Updated 13d ago

Evaluation Results

Method	Links
RETOOL-VIDEO 2026.05		81.5
Gemini-2.5-Flash-Lite 2026.05		78.5
Gemini 1.5 Pro 2026.05		74.3
Video-LLaMA3-7B 2026.05		73
InternVL3.5-30B-A3B 2026.05		73
Claude-Sonnet-4.5 2026.05		72.8
LLaVA-Video-7B 2026.04		71.4
LLaVA-Video-7B 2026.05		70.8
InternVL3-8B + LiteFrame 2026.05		70.7
Tango† 2026.04		70.4
LLaVA-Video 7B 2026.05		70.3
NVILA 2026.05		70.1
InternVL3-8B 2026.05		69.6
InternVL3-8B + FastVID 2026.05		69.3
GPT-5-Nano-2025-08-07 2026.05		69.2
InternVL-V2.5-8B 2026.05		68.9
Tango 2026.04		68.7
HoliTom 2026.04		68.2
FastVID 2026.04		68.1
VisionZip 2026.04		67.7
Tango† 2026.04		67.6
StreamAgent-7B 2025.08		67.2
Qwen2.5-VL-7B 2025.08		66.9
Tango 2026.04		66.7
InternVL3-8B + LiteFrame 2026.05		66.7
InternVL3-8B 2026.05		66.4
Flash-VStream 2026.05		66.3
HoliTom 2026.04		66.2
InternVL-V3.5-8B 2026.05		66.2
EvoStreaming-8B-InternVL-V3.5 2026.05		66
InternVL3-8B + LiteFrame 2026.05		65.7
DyCoke 2026.05		65.7
TimeChat-Online-7B 2025.08		65.4
LongVU-7B 2026.05		65.4
LLaVA-OV 7B 2026.05		65.2
Video-Thinker-7B 2026.05		65.2
FastVID 2026.04		65.1
InternVL3-8B + FastVID 2026.05		65.1
ParaVT-8B 2026.05		65
LLaVA-OneVision-7B 2026.05		64.7
GPT-4o 2026.05		64.6
GPT-4o 2026.05		64.6
DynaTok 2026.05		64.6
GPT-4o 2026.05		64.6
Qwen2.5-VL-7B 2026.04		64.4
VideoMind 2026.05		64.4
InternVL-V2-8B 2026.05		64
InternVL3-8B + FastVID 2026.05		64
FastV 2026.05		63.9
DynaTok 2026.05		63.2
VisionZip 2026.04		62.9
TimeChat-Online-7B 2026.05		62.9
DynaTok 2026.05		62.5
InternVL3-8B 2026.05		62.2
DyCoke 2026.05		62.1
DynaTok 2026.05		62.1
DynaTok 2026.05		62.1
VisionZip 2026.05		62.1
Dispider-7B 2025.08		61.7
Dispider-7B 2026.05		61.7
Video-R1-7B 2026.05		61.6
FastV 2026.05		61.5
FastV 2026.05		61.2
Tango 2026.04		61
Kangaroo-7B 2026.05		61
Kangaroo 2026.05		61
VisionZip 2026.04		60.6
Time-R1-7B 2026.05		60.5
DynaTok 2026.05		60.4
FastVID 2026.04		60.3
EvoStreaming-8B-Qwen2 2026.05		60.2
EvoStreaming-8B-Qwen2.5 2026.05		60.1
Qwen2-VL-7B 2026.05		60.1
ReWatch-R1-7B 2026.05		60.1
FastV 2026.05		59.8
VidCom 2026.04		59.7
LongVT-RFT-7B 2026.05		59.4
Conan-7B 2026.05		59.2
FastV 2026.05		58.9
Qwen2.5-VL-7B 2026.05		58.7
VideoChat-R1-7B 2026.05		58.7
VideoExplorer 2026.05		58.6
LLaVA-OneVision + LaVi 2025.07		58.5
VisionZip 2026.05		58.5
Qwen3-VL-8B 2026.05		58.3
FastV 2026.05		57.7
VisionZip 2026.05		57.1
DyCoke 2026.05		56.9
LongVA-7B 2025.08		56.3
LongVA-7B 2026.05		56.3
DyCoke 2026.05		55.8
VideoThinker 2026.05		54.8
VideoLLaMA2 + REGATE 2025.07		54.5
VisionZip 2026.05		54.4
FastV 2026.04		52.2
VisionZip 2026.05		52.2
VisionZip 2026.05		51.5
Flash-VStream-7B 2026.05		50.2
SAGE-7B 2026.05		49.7
Framemind 2026.05		48.6

Showing 100 of 114 rows