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

Mathematical Reasoning on IMO-Answer-Bench

86.3Accuracy

GPT-5.2

Updated 8d ago

Evaluation Results

Method	Links
GPT-5.2 2026.03		86.3
MSE+Distillation 2026.05		86.3
MSE+Distillation 2026.05		85.2
MSE+Distillation 2026.05		84
Kimi-K2.5 2026.03		81.8
Gemini-3-Pro 2026.03		81.3
TRICE-30B 2026.05		80.3
TRICE-30B 2026.05		80.3
Base 2026.05		79.9
GLM-4.7-Flash w/ recipe 2026.05		78.8
Intern-S1-Pro 2026.03		77.3
Nemotron-3-Nano-30B-A3B 2026.05		77
DeepSeek-V3.2-Thinking 2026.05		75
Qwen3.5-35B-A3B 2026.05		73.8
Qwen3-235B-A22B-Thinking 2026.05		73.8
Qwen3-VL-235B-Thinking 2026.03		72.3
TRICE-30B 2026.05		72.3
TRICE-4B 2026.05		68.9
GLM-4.7-Flash 2026.05		68.3
Qwen3-30B-A3B-Thinking-2507 2026.05		66.1
Qwen3.5-9B 2026.05		65
TRICE-4B 2026.05		61
GPT-OSS-20B 2026.05		60
Qwen3.5-4B 2026.05		59.5
Qwen3-Next-80B-A3B 2026.02		58
Qwen3-4B-Thinking-2507 2026.05		57
Qwen3-30B-A3B-2507 2026.02		54.34
Nanbeige4.1-3B 2026.02		53.38
Qwen3-30B-A3B-Instruct-2507 2026.05		52.3
Qwen3-4B-2507 2026.02		48
Qwen3-32B 2026.02		43.94
Nanbeige4-3B-2511 2026.02		38.25