Mathematical Reasoning on AIME 2025 (Avg@5)
92Avg@5 ScoreGLM-4.6
Evaluation results
| Method | Links | |
|---|---|---|
| GLM-4.6Evaluation Variant=seeds, Sample Count=302026.01 | 92 | |
| Gemini-3-Pro-PreviewEvaluation Variant=seeds, Sample Count=302026.01 | 90.7 | |
| DeepSeek-V3.2-thinkingEvaluation Variant=seeds, Sample Count=302026.01 | 90 | |
| GPT-5-highEvaluation Variant=seeds, Sample Count=302026.01 | 89.3 | |
| GPT-5.1-highEvaluation Variant=seeds, Sample Count=302026.01 | 88.7 | |
| GPT-5.1-highEvaluation Variant=VeRA-E, Sample Count=602026.01 | 88.3 | |
| DeepSeek-V3.2-thinkingEvaluation Variant=VeRA-E, Sample Count=602026.01 | 87.7 | |
| DeepSeek-V3.1-thinkingEvaluation Variant=seeds, Sample Count=302026.01 | 86.7 | |
| Gemini-3-Pro-PreviewEvaluation Variant=VeRA-E, Sample Count=602026.01 | 85.7 | |
| Gemini-2.5-ProEvaluation Variant=seeds, Sample Count=302026.01 | 85.3 | |
| GPT-5-highEvaluation Variant=VeRA-E, Sample Count=602026.01 | 85.3 | |
| DeepSeek-V3.1-thinkingEvaluation Variant=VeRA-E, Sample Count=602026.01 | 83 | |
| Seed-1.6-1015-highEvaluation Variant=seeds, Sample Count=302026.01 | 80.7 | |
| Seed-1.6-Thinking-0715Evaluation Variant=seeds, Sample Count=302026.01 | 80 | |
| Claude-Sonnet-4.5-thinkingEvaluation Variant=seeds, Sample Count=302026.01 | 78.7 | |
| Kimi-K2-thinkingEvaluation Variant=seeds, Sample Count=302026.01 | 78.7 | |
| Seed-1.6-Lite-1015-highEvaluation Variant=seeds, Sample Count=302026.01 | 77.3 | |
| Seed-1.6-1015-highEvaluation Variant=VeRA-E, Sample Count=602026.01 | 76.7 | |
| Minimax-M2Evaluation Variant=seeds, Sample Count=302026.01 | 76.7 | |
| GLM-4.6Evaluation Variant=VeRA-E, Sample Count=602026.01 | 76.3 | |
| Seed-1.6-Thinking-0715Evaluation Variant=VeRA-E, Sample Count=602026.01 | 75.3 | |
| qwen3-max-0923Evaluation Variant=seeds, Sample Count=302026.01 | 74.7 | |
| Gemini-2.5-ProEvaluation Variant=VeRA-E, Sample Count=602026.01 | 73 | |
| Kimi-K2-thinkingEvaluation Variant=VeRA-E, Sample Count=602026.01 | 71.7 | |
| Seed-1.6-Lite-1015-highEvaluation Variant=VeRA-E, Sample Count=602026.01 | 70 | |
| qwen3-max-0923Evaluation Variant=VeRA-E, Sample Count=602026.01 | 69.7 | |
| Claude-Sonnet-4.5-thinkingEvaluation Variant=VeRA-E, Sample Count=602026.01 | 68.7 | |
| Minimax-M2Evaluation Variant=VeRA-E, Sample Count=602026.01 | 57 | |
| GPT-5.1-chat-latestEvaluation Variant=seeds, Sample Count=302026.01 | 50.7 | |
| Kimi-K2-0905Evaluation Variant=seeds, Sample Count=302026.01 | 47.3 | |
| Kimi-K2-0905Evaluation Variant=VeRA-E, Sample Count=602026.01 | 44 | |
| GPT-5.1-chat-latestEvaluation Variant=VeRA-E, Sample Count=602026.01 | 41.7 |