Benchmark

HumanEval

Saturated. Retained because open-weight cards still report it.

Released
2021-07-07
Categories
coding
Openness
unknown
Source
model_reports
Reported scores
2

Reported scores

model_reports

ModelOrganizationReported valueReportedEvidence
Gemini 1.5 ProGoogle84.12024-03-08evidence
Qwen2.5-Coder-32B-InstructQwen92.72024-09-18evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.