A leak reveals that Anthropic is testing a more capable AI model "Claude Mythos"

Exclusive: Anthropic is testing ‘Mythos,’ its ‘most powerful AI model ever developed’ | Fortune

fortune.com

March 27, 2026

7 min read

🔥🔥🔥🔥🔥

45/100

Summary

Anthropic is testing a new AI model named 'Mythos,' which is claimed to be the most powerful model the company has developed to date. Early access customers are currently trialing this model, which represents a significant advancement in AI performance.

Key Takeaways

Anthropic is testing a new AI model called Claude Mythos, which is described as the most powerful AI model the company has ever developed.
The new model is believed to pose unprecedented cybersecurity risks and is currently being trialed by early access customers.
Anthropic is introducing a new tier of AI models named Capybara, which is larger and more capable than its previous Opus models.
Claude Mythos achieves significantly higher scores in software coding, academic reasoning, and cybersecurity compared to the previous best model, Claude Opus 4.6.

Read original article

View story →

Coverage & discussions

Hacker News · fortune.com

A leak reveals that Anthropic is testing a more capable AI model "Claude Mythos" ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Any alleged 'step change' over version 4.6 is worth paying attention to, indicating potential advancements in AI capabilities.
The Capybara tier concept could introduce meaningful improvements, suggesting that there may be significant innovations in this new model.

Concerns

The irony of discussing cybersecurity capabilities while leaking information highlights serious concerns about the model's security and reliability.
The cycle of declaring every new model as 'the one' is becoming tiresome, indicating skepticism about the actual advancements being made.