Mathematical Problem Solving on AIME VeRA-H Pro 2024-II
78.6Avg@5 AccuracyGPT-5-high
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GPT-5-high2026.01 | 78.6 | -10 | |
| GPT-5.1-high2026.01 | 74.3 | -20 | |
| DeepSeek-V3.2-thinking2026.01 | 68.6 | -22.9 | |
| GLM-4.62026.01 | 67.1 | -27.1 | |
| Seed-1.6-Thinking-07152026.01 | 67.1 | -20 | |
| Seed-1.6-1015-high2026.01 | 67.1 | -25.7 | |
| Gemini-2.5-Pro2026.01 | 65.7 | -27.1 | |
| Gemini-3-Pro-Preview2026.01 | 62.9 | -12.9 | |
| Claude-Sonnet-4.5-thinking2026.01 | 60 | -30 | |
| DeepSeek-V3.1-thinking2026.01 | 58.6 | -31.4 | |
| Seed-1.6-Lite-1015-high2026.01 | 57.1 | -31.4 | |
| Minimax-M22026.01 | 51.4 | -31.4 | |
| Qwen3-max-09232026.01 | 51.4 | -28.6 | |
| Kimi-K2-thinking2026.01 | 45.7 | -40 | |
| Kimi-K2-09052026.01 | 32.9 | -38.6 | |
| GPT-5.1-chat-latest2026.01 | 28.6 | -24.3 |