Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Code Generation on HumanEvalPlus (Accuracy)
Loading...
90.85
HumanEval+ Accuracy
Parallel-Synthesis
34.4196
49.0698
63.72
78.3702
Jun 12, 2026
HumanEval+ Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
HumanEval+ Accuracy
Parallel-Synthesis
Backbone=Qwen3-14B, Wo...
2026.06
90.85
Text-Serialization
Backbone=Qwen3-14B, Wo...
2026.06
90.24
Voting
Backbone=Qwen3-14B, Wo...
2026.06
89.63
KVLINK
Backbone=Qwen3-14B, Wo...
2026.06
87.8
Single
Backbone=Qwen3-14B, Wo...
2026.06
84.15
CacheBlend
Backbone=Qwen3-14B, Wo...
2026.06
51.22
APE
Backbone=Qwen3-14B, Wo...
2026.06
36.59
Feedback
Search any
task
Search any
task