ExploitBench ↗73.8%
Claude Mythos Preview (Early) · ExploitBench · Mean capability · % ↗
1 more reported result
70.6% · Claude Mythos Preview (Early) · ExploitBench · Mean capability · % · Epoch AI · ExploitBench ↗
Anthropic / MODEL EVIDENCE
Claude Mythos Preview (Early)
2 benchmarks · 4 observations · 2 sources · released 2026-04-07
Scores at a glance
2 benchmarks · large dot: Claude Mythos Preview (Early); small dots: closest models where reported
Claude Mythos Preview (Early) · METR task performance & time horizons · Average task accuracy ↗
1 more reported result
1044.78014 minutes · Claude Mythos Preview (Early) · METR task performance & time horizons · 50% success task horizon · METR-Horizon-v1.1 · Epoch AI · METR Autonomy / Task-Horizon Suite · public result data ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
Gemini 3.1 Pro Preview →Google
2 of 2 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
Names used in source tables
Claude Mythos Preview (Early)