Verilog Code Generation on RTLLM v1
72.9Pass@1CodeV-R1-7B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CodeV-R1-7BType=Verilog-Specific Reasoning Models2025.11 | 72.9 | 86.1 | — | |
| Deepseek-R1-671BType=Foundation Reasoning Models2025.11 | 64.8 | 82.9 | — | |
| QiMeng-CRUX-Final-7BType=Verilog-Specific General Models2025.11 | 62.8 | 69 | 71.5 | |
| Deepseek-V3-671BType=Foundation General Models2025.11 | 60.9 | 74.2 | — | |
| CodeV-R1-Distill-7BType=Verilog-Specific Reasoning Models2025.11 | 57.4 | 75.8 | — | |
| VeriPrefer-7BType=Verilog-Specific General Models2025.11 | 53.2 | 67.7 | — | |
| QWQ-32BType=Foundation Reasoning Models2025.11 | 50.9 | 70.6 | — | |
| OriGen-7BType=Verilog-Specific General Models2025.11 | 50.6 | 68.3 | 74.3 | |
| QiMeng-CRUX-SFT-7BType=Verilog-Specific General Models2025.11 | 49 | 64.1 | 72 | |
| Qwen2.5-Coder-32BType=General Code Models2025.11 | 47.9 | 67.7 | — | |
| GPT-4oType=Foundation General Models2025.11 | 41.7 | 65.9 | — | |
| CodeV-Qwen2.5-7BType=Verilog-Specific General Models2025.11 | 39.3 | 63.5 | 74.2 | |
| CodeV-Qwen1.5-7BType=Verilog-Specific General Models2025.11 | 36.6 | 53.3 | 61.3 | |
| RTLCoder-6.7BType=Verilog-Specific General Models2025.11 | 35.8 | 40.3 | 43.1 | |
| Qwen2.5-Coder-7BType=General Code Models2025.11 | 32.2 | 48.2 | 56 | |
| Deepseek-Coder-6.7BType=General Code Models2025.11 | 30.2 | 44 | 52.5 |