Benchmark

DRACO

Deep-research style long-context agent tasks; the retrieval and tool stack is part of the score.

Categories
long_context
Openness
unknown
Source
model_reports
Reported scores
1

Reported scores

model_reports

ModelOrganizationReported valueReportedEvidence
Hy4 previewTencent77.22026-08-28evidence

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.