Math Evaluation on Math ID
97.26AccuracyGPT-5.2-chat-latest
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-5.2-chat-latest2026.05 | 97.26 | 80.53 | |
| GPT-5-chat-latest2026.05 | 96.94 | 75.22 | |
| GSPO+AIR2026.05 | 94.93 | 68.14 | |
| GRPO2026.05 | 93.81 | 64.6 | |
| GRPO+AIR2026.05 | 93.64 | 63.72 | |
| GSPO2026.05 | 93.56 | 65.49 | |
| GRPO+V-REx2026.05 | 93.02 | 60.71 | |
| Llama-4-Maverick2026.05 | 92.84 | 61.06 | |
| GSPO+V-REx2026.05 | 92.77 | 58.94 | |
| Qwen-3-14B2026.05 | 92.28 | 58.41 | |
| Gemini-3-Pro-Preview2026.05 | 90.02 | 33.63 | |
| Gemini-2.5-Pro2026.05 | 89.62 | 37.17 | |
| Qwen-2.5-14B-Instruct2026.05 | 88.99 | 48.67 |