reasoning benchmark
HumanEval-Average
A variant of the HumanEval benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
HumanEval-Average Ranking
Higher score ranks better on this benchmark.
| 1 | MA | 61.5% | 100.0% | 1 | C |
HumanEval-Average Highlights
The leading models and scores on this benchmark.
What is HumanEval-Average?
What HumanEval-Average measures and how its scores work.
A variant of the HumanEval benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- HumanEval-Average
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- humaneval-average|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about HumanEval-Average.
Which model scores highest on HumanEval-Average?
Codestral-22B is currently ranked first with 61.5%.
What does HumanEval-Average measure?
A variant of the HumanEval benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.