language benchmark
MultiPL-E
MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.
Updated Aug 7, 2026
Published models13
Registry coverage13
MetricScore
EvidenceB
MultiPL-E leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 87.9% | 100.0% | 13 | C | |
| 2 | AC | 87.8% | 91.7% | 13 | C | |
| 3 | AC | 86.1% | 83.3% | 13 | C | |
| 4 | MA | 85.7% | 75.0% | 13 | C | |
| 5 | MA | 85.7% | 66.7% | 13 | C | |
| 6 | AC | 75.4% | 58.3% | 13 | C | |
| 7 | AC | 75.1% | 50.0% | 13 | C | |
| 8 | AC | 72.8% | 41.7% | 13 | C | |
| 9 | AC | 70.4% | 33.3% | 13 | C | |
| 10 | AC | 69.2% | 25.0% | 13 | C | |
| 11 | AC | 65.9% | 16.7% | 13 | C | |
| 12 | AC | 65.8% | 8.3% | 13 | C | |
| 13 | AC | 59.1% | 0.0% | 13 | C |
Score distribution
Top published rows on the benchmark's original scale.
MultiPL-E highlights
The top published results on this benchmark's own scale.
What is MultiPL-E?
Definition and scoring fields from the benchmark registry.
MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MultiPL-E
- Modality
- text
- Primary category
- language
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- multipl-e|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MultiPL-E.
Which model scores highest on MultiPL-E?
Qwen3-235B-A22B-Instruct-2507 is currently ranked first with 87.9%.
What does MultiPL-E measure?
MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
13 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is marked as eligible for the current LLMBoard capability methodology.