Benchmark

DROP Benchmark

What is DROP?

Legacy reading-comprehension baseline. Benchmark Radar tracks 3 reported scores for DROP, cited by 6 source documents.

Released
2019-03-01
Evaluates
reasoning
Openness
unknown
Importer
Model reports
Reported scores
3

Source provenance

DROP links

DROP results and reported scores

Model reports

ModelOrganizationReported valueReportedEvidence
DeepSeek-R1DeepSeek92.22025-01-22evidence
DeepSeek-V3DeepSeek91.62024-12-27evidence
Gemini 1.5 ProGoogle74.92024-03-08evidence

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.

Which sources cite DROP?

Related reasoning benchmarks