reasoning benchmark
EvalPlus
A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases
Updated Aug 7, 2026
Published models4
Registry coverage4
MetricScore
EvidenceB
EvalPlus leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MA | 0.803 points | 100.0% | 4 | C | |
| 2 | AC | 0.79 points | 66.7% | 4 | C | |
| 3 | AC | 0.776 points | 33.3% | 4 | C | |
| 4 | AC | 0.703 points | 0.0% | 4 | C |
Score distribution
Top published rows on the benchmark's original scale.
EvalPlus highlights
The top published results on this benchmark's own scale.
What is EvalPlus?
Definition and scoring fields from the benchmark registry.
A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases
Scores are shown in points. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- EvalPlus
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- evalplus|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about EvalPlus.
Which model scores highest on EvalPlus?
Kimi K2 Base is currently ranked first with 0.803 points.
What does EvalPlus measure?
A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
4 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.