List your company

Mathematical Reasoning on IMO-Answer-Bench

86.3AccuracyGPT-5.2
36.32849.301562.27575.2485Feb 13, 2026Mar 1, 2026Mar 18, 2026Apr 4, 2026Apr 21, 2026May 8, 2026May 25, 2026

Evaluation results

MethodLinks
GPT-5.22026.0386.3
MSE+DistillationModel=Qwen3.5-27B, Batch=32026.0586.3
MSE+DistillationModel=Qwen3.5-27B, Batch=22026.0585.2
MSE+DistillationModel=Qwen3.5-27B, Batch=12026.0584
Kimi-K2.5Number of Parameters=1T-A32B2026.0381.8
Gemini-3-Pro2026.0381.3
TRICE-30BTool Usage=true, Parameter Scale=~30B, Evaluation Protocol=unified2026.0580.3
TRICE-30BTool Use=Yes2026.0580.3
BaseModel=Qwen3.5-27B2026.0579.9
GLM-4.7-Flash w/ recipeTool Usage=true, Parameter Scale=~30B, Evaluation Protocol=unified2026.0578.8
Intern-S1-ProNumber of Parameters=1T-A22B2026.0377.3
Nemotron-3-Nano-30B-A3BTool Usage=true, Parameter Scale=~30B, Evaluation Protocol=unified2026.0577
DeepSeek-V3.2-ThinkingTool Use=No2026.0575
Qwen3.5-35B-A3BTool Usage=false, Parameter Scale=~30B, Evaluation Protocol=unified2026.0573.8
Qwen3-235B-A22B-ThinkingTool Use=No2026.0573.8
Qwen3-VL-235B-ThinkingNumber of Parameters=235B-A22B, Thinking Configuration=true2026.0372.3
TRICE-30BTool Usage=false, Parameter Scale=~30B, Evaluation Protocol=unified2026.0572.3
TRICE-4BTool Usage=true, Parameter Scale=<10B, Evaluation Protocol=unified2026.0568.9
GLM-4.7-FlashTool Usage=true, Parameter Scale=~30B, Evaluation Protocol=unified2026.0568.3
Qwen3-30B-A3B-Thinking-2507Tool Usage=false, Parameter Scale=~30B, Evaluation Protocol=unified2026.0566.1
Qwen3.5-9BTool Usage=false, Parameter Scale=<10B, Evaluation Protocol=unified2026.0565
TRICE-4BTool Usage=false, Parameter Scale=<10B, Evaluation Protocol=unified2026.0561
GPT-OSS-20BTool Usage=true, Parameter Scale=~30B, Evaluation Protocol=unified2026.0560
Qwen3.5-4BTool Usage=false, Parameter Scale=<10B, Evaluation Protocol=unified2026.0559.5
Qwen3-Next-80B-A3BParameters=80B2026.0258
Qwen3-4B-Thinking-2507Tool Usage=false, Parameter Scale=<10B, Evaluation Protocol=unified2026.0557
Qwen3-30B-A3B-2507Parameters=30B2026.0254.34
Nanbeige4.1-3BParameters=3B2026.0253.38
Qwen3-30B-A3B-Instruct-2507Tool Usage=false, Parameter Scale=~30B, Evaluation Protocol=unified2026.0552.3
Qwen3-4B-2507Parameters=4B2026.0248
Qwen3-32BParameters=32B2026.0243.94
Nanbeige4-3B-2511Parameters=3B2026.0238.25