Verilog Code Generation on VerilogEval 2.0 (test) benchmark leaderboard · SOTA2 Research

Verilog Code Generation on VerilogEval 2.0 (test)

82.47Pass@1VeriRL-CodeQwen2.5 + VERIMOA

Evaluation Results

Method	Links
VeriRL-CodeQwen2.5 + VERIMOA2025.10		82.47	87.91	90.68
Qwen2.5-Coder-32B + VERIMOA2025.10		73.31	79.05	81.27
Qwen2.5-Coder-14B + VERIMOA2025.10		66.86	73.24	76.38
VeriRL-CodeQwen2.52025.10		66.28	71.35	73.43
VeriRL-DeepSeek-Coder2025.10		64.57	68.96	71.78
Qwen2.5-Coder-7B + VERIMOA2025.10		60.96	68.13	70.69
HaVen-CodeQwen1.52025.10		55.4	59.73	62.76
OriGen-DeepSeek-Coder-7B-v1.52025.10		51.19	54.88	56.78
RTLCoder-DeepSeek-Coder2025.10		39.67	43.66	45.99
RTLCoder-Mistral2025.10		35.62	36.63	37.71