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

Geometric Reasoning on GeoAux-Bench-Core Olympiad Senior

89.45Accuracy

Qwen3-VL-235B-Thk.

Updated 1mo ago

Evaluation Results

Method	Links
Qwen3-VL-235B-Thk. 2026.03		89.45
Qwen3-VL-30B-Thk. 2026.03		83.97
Gemini-2.5-Pro 2026.03		82.28
GPT-5 2026.03		79.32
Qwen3-VL-235B-Ins. 2026.03		78.9
Qwen3-VL-30B-Ins. 2026.03		75.53
Doubao-seed-1.6 2026.03		72.57
GLM-4.5V 2026.03		72.15
MiMo-VL-7B-SFT 2026.03		71.73
MiMo-VL-7B-RL 2026.03		71.73
InternVL3-8B 2026.03		67.93
Qwen2.5-VL-Ins. 2026.03		65.82
Gemini-2.5-Flash 2026.03		64.56
Claude-Opus-4.1 2026.03		58.23
GPT-4o 2026.03		57.38
GLM-4.1V-Thk. 2026.03		56.12
Claude-Sonnet-4.5 2026.03		55.7
InternVL3.5-8B 2026.03		54.85
MathCanvas-7B 2026.03		46.84
BAGEL-7B-MoT 2026.03		34.6
BAGEL-Zebra-CoT 2026.03		31.65