Model

GenClaw

Vendor N/A

Agentic / Tool Use

Agentic / Tool Use
BenchmarkScoreMetricHarnessSourceConfidenceLink
Savanna36.5scoregenclaw ∗ claude-opus-5Independentmediumarxiv.org
Savanna34.2scoregenclaw claude-opus-4.6Independentmediumarxiv.org

Reproduced by the community

Loading community runs…