ExploitBench (Cap%)

AI benchmark results and model performance

Type: percentageCybersecurity

78%

78% - 78%

78% - Claude Fable 5

Scores Over Time

Individual benchmark scores plotted by date.

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Anthropic	Claude Fable 5	09 Jun 2026	78%	Cap%; starred row in Anthropic's launch table	Yes	Source
Anthropic	Claude Mythos 5	09 Jun 2026	78%	Cap%; starred row in Anthropic's launch table	Yes	Source