// BENCHMARK

MultiPL-E benchmark

AI model leaderboard for the MultiPL-E benchmark. Compare how large language models score on MultiPL-E, see the full ranking, and understand what this AI benchmark measures. Ling-1T currently leads with 77.91. Multi-language code-generation benchmark — translates HumanEval/MBPP into 18+ programming languages.

Leaderboard

# Model Organization Score Variant Source
#1 Ling-1T Ant Group 77.91 pass@1 official ↗
#2 DeepSeek-V3.1-Terminus DeepSeek 77.68 non-thinking-cited-ling-1t official ↗
#3 GPT-5 OpenAI 76.66 main-cited-ling-1t official ↗
#4 Kimi K2 Moonshot 73.54 instruct-0905-cited-ling-1t official ↗
#5 gemini-2.5-pro Google DeepMind 71.48 lowthink-cited-ling-1t official ↗
#6 Granite-4.0-H-Small IBM 57.37 pass@1-multiple official ↗
#7 Granite-4.0-H-Tiny IBM 55.83 pass@1-multiple official ↗
#8 Qwen2-57B-A14B Qwen 49.8 base-from-hf-readme official ↗
#9 Granite-4.0-H-Micro IBM 49.46 pass@1-multiple official ↗
#10 Qwen2-7B Qwen 46.3 base-from-hf-readme official ↗
#11 Yi-1.5-34B 01.AI 39.5 cited-qwen2-57b official ↗
#12 Mixtral 8x7B Mistral AI 39 cited-qwen2-57b official ↗
#13 Qwen1.5-32B Qwen 38.5 cited-qwen2-57b official ↗

Frequently asked questions about MultiPL-E

What is the MultiPL-E benchmark?

Multi-language code-generation benchmark — translates HumanEval/MBPP into 18+ programming languages.

How is the MultiPL-E benchmark scored?

MultiPL-E is scored using the pass@1 metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.

Which AI model scores highest on MultiPL-E?

As of the latest reported scores on GenAIList, Ling-1T achieves the highest result on MultiPL-E with a score of 77.91.

Is a higher MultiPL-E score better?

Yes. On MultiPL-E a higher score indicates better performance, so models near the top of the leaderboard are the strongest.