Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multilingual Code Generation on MultiPL-E (pass@1)
Loading...
91.2
Pass@1 Score
Gemini-3-Pro
66.344
72.797
79.25
85.703
Jun 16, 2026
Pass@1 Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Pass@1 Score
Gemini-3-Pro
2026.06
91.2
Claude-Opus-4.5
2026.06
91
Qwen3-235B-A22B-Instruct-2507
2026.06
87.9
Qwen3-Coder-480B-A35B-Instruct
2026.06
87.4
GPT-5.1
2026.06
86.2
DeepSeek-V3.2
2026.06
85.8
Kimi-K2-Instruct-0905
2026.06
85.7
Kimi-Dev-72B
2026.06
80.3
Qwen2.5-Coder-32B-Instruct
2026.06
79.6
Qwen2.5-Coder-14B-Instruct
2026.06
78.8
Qwen2.5-Coder-7B-Instruct
2026.06
75.4
Seed-Coder-8B-Instruct
2026.06
75.1
LoopCoder-v2
R (number of refinemen...
2026.06
73.9
DeepSeek-Coder-V2-Lite-Instruct
2026.06
71.5
LoopCoder-v2
R (number of refinemen...
2026.06
69.8
LoopCoder-v2
R (number of refinemen...
2026.06
69.5
GLM-4.7
2026.06
69
LoopCoder-v2
R (number of refinemen...
2026.06
67.3
Feedback
Search any
task
Search any
task