claude-mythos-restricted-access-model

IN premise — entries/2026/06/21/wiki-Claude_language_model.md

Created 2026-06-21T09:50:09+00:00

Claude Mythos is a separate, more capable model line released under restricted access (Project Glasswing) for cybersecurity vulnerability discovery, not generally available to the public.

Summary

There is a more advanced version of Claude that exists outside the public-facing product, built specifically to hunt down security flaws in software and infrastructure. This matters because it means a tier of AI reasoning capability exists in the wild that is purpose-built to find vulnerabilities, which changes the threat model for any system that could be probed by it.

Dependents

These beliefs depend on this one:

IN claude-tiered-strategy-manages-capability-safety-spectrum — Claude's model strategy creates a structured capability-safety spectrum: three public tiers (Haiku/Sonnet/Opus) with safety classification scaling by capability (Opus 4 at Level 3), plus a restricted tier (Mythos) for the highest-capability models — systematically linking access to risk.
OUT fable-5-provides-public-mythos-class-access — Fable 5 provides public access to Mythos-class model capability, serving as a safety-guardrailed version of the restricted Mythos line.