Claude Mythos 5

Model Overview

Claude Mythos 5 is a highly restricted, enterprise-grade frontier model released by Anthropic on June 9, 2026. It shares the same core architecture as Claude Fable 5 but contains advanced capabilities specifically optimized for life sciences, biology research, and high-security cyber contexts. It is a "closed-source" model exclusively available through Project Glasswing to vetted partners.

Capabilities

• Unrestricted Dual-Use Performance: Operates without the standard safety classifiers found in general-release models like Fable 5, allowing it to perform deep, unrestricted work in sensitive areas.
• Advanced Biological Modeling: Highly optimized for biological research and life sciences.
• Cybersecurity & Vulnerability Discovery: Capable of autonomously identifying and patching high-severity vulnerabilities in critical software infrastructure.
• Massive Context: Supports a 1 million token context window across text, image, code, and multimodal inputs.

Example Use Cases

• Defensive cybersecurity research and infrastructure patching.
• High-stakes life sciences and biodefense research.
• Extremely advanced, sensitive knowledge work limited to government agencies, security firms, and top-tier research organizations.

Performance & Benchmarks

Claude Mythos 5 represents the absolute pinnacle of Anthropic's capabilities as of mid-2026. Its performance in vulnerability discovery and complex domain reasoning is unparalleled, bypassing the fallback mechanisms that constrain Fable 5. Detailed benchmark figures remain undisclosed due to the restricted nature of the model.

Intended Use & Limitations

Restricted Availability: Claude Mythos 5 is not available to the general public. Access is strictly managed through Project Glasswing, a collaborative initiative involving major technology companies and government agencies. In June 2026, the model briefly faced suspension due to US government directives regarding export controls, highlighting the sensitive and potentially dangerous nature of its dual-use capabilities.

About Anthropic

Anthropic is an AI safety and research company that develops the Claude family of models. With initiatives like Project Glasswing and the release of highly guarded models like Mythos 5, Anthropic focuses on managing the responsible deployment of powerful frontier AI while prioritizing security, alignment, and safety.