List your company

Code Generation on HumanEval (M, Speedup)

16.47Functional Score MDOUBLE
0.85964.91238.96513.0177Jan 9, 2026

Evaluation results

MethodLinks
DOUBLEBase Model=Deepseek-33B2026.0116.474.45
PEARLBase Model=Deepseek-33B2026.0114.513.49
DOUBLEBase Model=LLaMA3.3-70B2026.0114.355.33
DOUBLEBase Model=Qwen3-14B2026.0112.651.65
PEARLBase Model=LLaMA3.3-70B2026.0110.694.39
DOUBLEBase Model=Qwen3-32B2026.0110.092.84
DOUBLEBase Model=LLaMA2-70B2026.018.653.42
PEARLBase Model=Qwen3-32B2026.017.782.35
PEARLBase Model=LLaMA2-70B2026.017.243.09
PEARLBase Model=Qwen3-14B2026.017.241.38
EAGLE3Base Model=LLaMA3.3-70B2026.016.514.56
OuroborosBase Model=Deepseek-33B2026.016.383.15
SpsBase Model=LLaMA2-70B2026.015.422.58
SpsBase Model=LLaMA3.3-70B2026.015.43.53
SpsBase Model=Deepseek-33B2026.015.092.15
OuroborosBase Model=LLaMA2-70B2026.015.082.08
SpsBase Model=Qwen3-32B2026.014.271.53
SpsBase Model=Qwen3-14B2026.013.371.26
EAGLE3Base Model=Qwen3-14B2026.012.991.91
TRBase Model=Deepseek-33B2026.012.832.14
EAGLE3Base Model=Qwen3-32B2026.012.832.27
TRBase Model=LLaMA2-70B2026.012.432.02
LookaheadBase Model=Deepseek-33B2026.012.341.75
PLDBase Model=Deepseek-33B2026.011.861.68
LookaheadBase Model=LLaMA2-70B2026.011.861.58
PLDBase Model=LLaMA3.3-70B2026.011.761.58
PLDBase Model=LLaMA2-70B2026.011.661.53
PLDBase Model=Qwen3-14B2026.011.521.38
PLDBase Model=Qwen3-32B2026.011.461.16