Code Generation on HumanEval+ ko
92.1ScoreGLM-4.6
Evaluation Results
| Method | Links | |
|---|---|---|
| GLM-4.6Parameters=357B-A32B, Non-Thinking Mode=true2026.01 | 92.1 | |
| DeepSeek-V3.1Parameters=685B-A37B, Non-Thinking Mode=true2026.01 | 86.6 | |
| A.X K1Parameters=519B-A33B, Non-Thinking Mode=true2026.01 | 75.6 |