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

Visual Mathematical Reasoning on DynaMath

81.42Accuracy

Gemini2.5-Pro

Updated 1mo ago

Evaluation Results

Method	Links
Gemini2.5-Pro 2026.03		81.42
CodePercept-32B-S1 2026.03		79.5
CodePercept-32B-S1 2026.03		77.41
Qwen3-VL-235A22B-Instruct 2026.03		77.39
Qwen3-VL-32B-Instruct 2026.03		75.78
Qwen3-VL-32B-Instruct 2026.03		75.54
CodePercept-8B-S1 2026.03		75.05
Qwen3-VL-8B-Instruct 2026.03		73.69
Claude-Opus 4.1-Thinking 2026.03		73.25
CodePercept-8B-S1 2026.03		73.2
CodePercept-4B-S1 2026.03		72.4
Qwen3-VL-8B-Instruct 2026.03		72.19
Qwen3-VL-30A3B-Instruct 2026.03		71.67
CodePercept-4B-S1 2026.03		71.38
Qwen3-VL-4B-Instruct 2026.03		71.22
GPT5-Thinking 2026.03		71
Qwen3-VL-4B-Instruct 2026.03		69.4
Qwen2.5-VL-72B 2026.03		68.28
InternVL3.5-8B 2026.03		68.12
GLM-4.1V-9B 2026.03		66.17
MiniCPM-V-4.5 2026.03		65.44
Intern-S1-8B 2026.03		63.61
KeyeVL1.5-8B 2026.03		62.37
GPT-5 2026.01		60.9
Active-Zero 2026.02		59.04
VisPlay 2026.02		58.98
VisionZero-CLEVR 2026.02		58.7
VisionZero-RealWorld 2026.02		58.62
VisionZero-Chart 2026.02		58.4
Base Model 2026.02		58.26
EvoLMM 2026.02		58.06
Gemini-2.5-Pro 2026.01		56.3
GLM-4.5V 2026.01		53.9
Active-Zero 2026.02		51.24
VisPlay 2026.02		50.02
Base Model 2026.02		46.33
COGFLOW 2026.01		46.2
Doubao-1.5-Pro 2026.01		44.9
GLM-4.1V 2026.01		42.5
Claude-3.7-Sonnet 2026.01		39.7
InternVL3.5 2026.01		37.7
Keye-VL 2026.01		37.3
InternVL3 2026.01		25.5
Qwen2.5-VL 2026.01		21
Ovis 2026.01		20.4