Anthropic Launches Its Most Powerful Model Claude Mythos, Outperforming Opus 4.6

Anthropic launches the preview version of its most powerful model, Claude Mythos, outperforming Opus 4.6. The model can discover thousands of high-risk vulnerabilities, with attack capabilities exceeding the vast majority of hackers. Due to excessive security risks, the model is not yet open to the public. It is only available through "Project Glasswing" to giants like Amazon, Apple, and Google, as well as over 40 infrastructure companies, for defensive security hardening.

Claude is an artificial intelligence, trained by Anthropic using Constitutional AI to be safe, accurate, and secure — the trusted assistant for you to do your best work.

April 9, 2025 • This article tells the story of several founders who left OpenAI, bringing with them the concept that "AI must be ethical," to establish the company Anthropic, and cultivated (as they say) "the most upright citizen in the AI world," Claude.

June 9, 2026 Anthropic released Claude Fable 5, positioned as the public safe version of the Mythos-level model. The official statement claims that in almost all bas...

2 days ago • Claude is a large language model family developed by the American AI company Anthropic. The first version was released on March 15, 2023. The founding team includes former core members of OpenAI and GPT-3 developers. The model uses the Constitutional AI training method, support...

This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.