general benchmark
Instruct HumanEval
Instruction-based variant of HumanEval benchmark for evaluating large language models' code generation capabilities with functional correctness using pass@k metric on programming problems
Updated Aug 7, 2026
Published models1
Registry coverage1
MetricScore
EvidenceB
Instruct HumanEval leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | NV | 73.8% | 100.0% | 1 | C |
Instruct HumanEval highlights
The top published results on this benchmark's own scale.
What is Instruct HumanEval?
Definition and scoring fields from the benchmark registry.
Instruction-based variant of HumanEval benchmark for evaluating large language models' code generation capabilities with functional correctness using pass@k metric on programming problems
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- Instruct HumanEval
- Modality
- text
- Primary category
- general
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- instruct-humaneval|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about Instruct HumanEval.
Which model scores highest on Instruct HumanEval?
Llama 3.1 Nemotron 70B Instruct is currently ranked first with 73.8%.
What does Instruct HumanEval measure?
Instruction-based variant of HumanEval benchmark for evaluating large language models' code generation capabilities with functional correctness using pass@k metric on programming problems
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 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.