List your company

Math on Math OOD

92.04AccuracyGPT-5.2-chat-latest
78.686482.153285.6289.0868May 20, 2026

Evaluation results

MethodLinks
GPT-5.2-chat-latest2026.0592.0488.5
GPT-5-chat-latest2026.0590.2771.68
Qwen-3-14B2026.0588.562.83
GRPO+AIR2026.0588.4961.06
GSPO2026.0587.6154.87
GSPO+AIR2026.0586.2851.33
Llama-4-Maverick2026.0584.5146.02
GSPO+V-REx2026.0584.0745.13
Gemini-3-Pro-Preview2026.0583.6343.36
GRPO+V-REx2026.0583.1942.48
GRPO2026.0582.340.71
Qwen-2.5-14B-Instruct2026.0581.8625.66
Gemini-2.5-Pro2026.0579.220.35