Claude Mythos 5.1

Why teams choose it

Tradeoffs to know

• Not on Claude Pro, Max, Team, or general Enterprise self-serve plans.
• Biology and cyber permissions remain program-gated even after an invite.

When not to use this

• Current trusted-access rollout is limited; do not catalog it as a drop-in Fable 5.1 replacement.
• Not ideal for simple tasks where cheaper models in the same lineup are good enough.
• Avoid for regulated or high-stakes outputs without evaluations that mimic your tooling, data, and review process.

Technical specs

Inputs

text, image

Outputs

text

Capabilities

defensive cybersecurity, life-science research, long-horizon coding, adaptive thinking, long context

License

Proprietary API (invite only)

Model string

claude-mythos-5-1

Benchmarks

{
  "source": "https://www.anthropic.com/claude-fable-and-mythos-5-1",
  "same_specs_as": "claude-fable-5-1",
  "vendor_reported": true,
  "terminal_bench_4_0_mythos": 60.9
}