Video Knowledge Acquisition on Video-MMLU benchmark leaderboard · SOTA2 Research

Video Knowledge Acquisition on Video-MMLU

0.674Notebook ScoreClaude-3.5-sonnet

Evaluation Results

Method	Links
Claude-3.5-sonnet2026.02		0.674	0.712	0.693
GPT-4o2026.02		0.539	0.449	0.494
Qwen2.5-VL-7B2026.02		0.42	0.329	0.375
Demo-ICL2026.02		0.41	0.504	0.457
Gemini-1.5-Flash2026.02		0.395	0.478	0.436
LLaVA-NeXT-Video-7B2026.02		0.358	0.274	0.316
LLaVA-OneVision-OV2026.02		0.346	0.334	0.34
InternVL2.5-8B2026.02		0.345	0.445	0.395
Qwen2-VL-7B2026.02		0.342	0.234	0.288
InternVL2-8B2026.02		0.314	0.167	0.241
Qwen-VL2026.02		0.244	0.196	0.22
mPLUG-Owl32026.02		0.226	0.226	0.226
LLaVA-1.52026.02		0.223	0.091	0.157
LLaVA-NeXT (Mistral)2026.02		0.203	0.315	0.258
Cambrian-8B2026.02		0.202	0.052	0.127
InstructBLIP-7B2026.02		0.193	0.029	0.111
LLaVA-NeXT (Vicuna)2026.02		0.181	0.249	0.215
LLaVA-NeXT (LLaMA3)2026.02		0.165	0.301	0.233
Video-LlaVA-7B2026.02		0.153	0.165	0.159