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

Long-document Visual Question Answering on MMLongBench Overall

90.77Average Score

GPT-5.5

Updated 20d ago

Evaluation Results

Method	Links
GPT-5.5 2026.05		90.77
Gemini-3.1-Pro 2026.05		83.66
Gemini-3.1-Flash 2026.05		80.13
Gemini-2.5-Pro 2026.05		74.37
GPT-5.4 2026.05		69.41
Gemma4-31B 2026.05		67.1
Qwen2.5-VL-72B 2026.05		60.83
Qwen2.5-VL-32B 2026.05		58.31
MMProLong 2026.05		57.7
Gemma3-27B 2026.05		52.63
Qwen2.5-VL-7B 2026.05		50.59
InternVL3-38B 2026.05		48.88
Gemma3-12B 2026.05		47.76
Gemma4-26B-A4B 2026.05		47.74
InternVL3-14B 2026.05		47.47
InternVL3-8B 2026.05		47.15
Gemma3-4B 2026.05		38.03
Gemma4-E4B 2026.05		37.08
Gemma4-E2B 2026.05		34.24
InternVL3.5-38B 2026.05		33.81
InternVL3.5-14B 2026.05		30.23
InternVL3.5-8B 2026.05		27.56