List your company
ResearchFormerly WizWand

Video Understanding on Video-MME w/o sub.

76.6AccuracyRETOOL-VIDEO
42.2851.1960.169.01May 13, 2026May 17, 2026May 22, 2026May 27, 2026Jun 1, 2026Jun 6, 2026Jun 11, 2026
Updated 4mo ago

Evaluation results

MethodLinks
RETOOL-VIDEOSize=9B2026.0576.6
Claude-Sonnet-4.52026.0575.3
Gemini 1.5 ProSetting=best-setting numbers from official reports2026.0575
Gemini-1.5-Pro# Params=-2026.0675
Gemini-2.5-Flash-Lite2026.0572.7
GPT-4o2026.0571.9
GPT-4oSetting=best-setting numbers from official reports2026.0571.9
GPT-4o# Params=-2026.0671.9
Gemini-1.5-Flash# Params=-2026.0670.3
VideoSeek2026.0570.1
STAR2026.0570
InternVL3.5-30B-A3BSize=30B2026.0568.7
DVD2026.0567.3
GPT-5-Nano-2025-08-072026.0566.2
NVILASize=8B2026.0564.2
ParaVT-8BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0562.1
Video-Thinker-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0561.9
Flash-VStreamSize=7B2026.0561.2
FramemindSize=7B2026.0560.9
GPT4-V2026.0560.7
VILA# Params=40B2026.0660.1
HYDRA-X# Params=7B2026.0660
Qwen3-VL-8BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0559.9
GPT-4V# Params=-2026.0659.9
LongVT-RFT-7BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0559.5
Time-R1-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0558.9
ReWatch-R1-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0558.8
VideoRFT-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0558.5
VideoMindSize=7B2026.0558.2
LLaVA-OV# Params=7B2026.0658.2
Video-R1-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0557.6
Show-o2# Params=7B2026.0657.4
KangarooSize=8B2026.0556
IXC-2.5# Params=7B2026.0655.8
Qwen2.5-VL-7BSetting=direct-answer setting (native single-pass prompt)2026.0555.7
Conan-7BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0555.5
VideoThinkerSize=7B2026.0553.7
LongVA# Params=7B2026.0652.6
VideoChat-R1-7BSetting=reasoning-enhanced setting (<think>→<answer>)2026.0550.4
TUNA# Params=1.5B2026.0649.1
Show-o2# Params=1.5B2026.0648
VideoLLaMA2# Params=7B2026.0647.9
VideoLLaMA2Size=8×7B2026.0546.6
VideoZoomer-7BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0545.3
SAGE-7BSetting=tool-augmented setting (<think>→<tool_call>→<answer>)2026.0544.1
ShareGPT4VideoSize=8B2026.0543.6