List your company

Geometric Reasoning on GeoAux-Bench-Core Olympiad Junior

95.17AccuracyQwen3-VL-235B-Thk.
32.87449.04765.2281.393Mar 19, 2026

Evaluation results

MethodLinks
Qwen3-VL-235B-Thk.Think=true2026.0395.17
Qwen3-VL-30B-Thk.Think=true2026.0393.24
Qwen3-VL-235B-Ins.Think=false2026.0391.79
Qwen3-VL-30B-Ins.Think=false2026.0389.37
GPT-5Think=true2026.0388.41
Gemini-2.5-ProThink=true2026.0388.24
Doubao-seed-1.6Think=true2026.0386.47
GLM-4.5VThink=false2026.0382.61
Gemini-2.5-FlashThink=true2026.0382.13
MiMo-VL-7B-SFTThink=false2026.0380.68
MiMo-VL-7B-RLThink=true2026.0380.68
InternVL3.5-8BThink=false2026.0366.67
Claude-Sonnet-4.5Think=true2026.0360.39
Claude-Opus-4.1Think=true2026.0358.45
InternVL3-8BThink=false2026.0355.56
GLM-4.1V-Thk.Think=true2026.0353.14
Qwen2.5-VL-Ins.Think=false2026.0352.17
MathCanvas-7BThink=false2026.0351.69
GPT-4oThink=false2026.0342.51
BAGEL-7B-MoTThink=false2026.0338.16
BAGEL-Zebra-CoTThink=false2026.0335.27