Mythos
A frontier-class model in the Claude family with domain-specific safety safeguards removed, reserved for specialized restricted domains.
What it is
Mythos is a specialized version of Anthropic's Fable model. It shares the same underlying architecture and reasoning capabilities as Fable but has specific safety restrictions lifted in certain domains (e.g. cybersecurity). Access to Mythos is highly restricted to authorized entities (such as cybersecurity partners and Project Glasswing).
When you would use it
You use Mythos only when authorized and working within a specialized domain (like cybersecurity research) that requires the removal of standard safety restrictions to perform analysis.
Common operations
- Performing advanced penetration testing and security vulnerability analysis.
- Researching threat intelligence and simulating adversary techniques in a sandbox environment.
Related terms
Where this is taught
No learning path uses this term yet. Browse the Learning Atlas for guided sequences through related ideas.