Anthropic rolls out Claude Fable 5.1 with cheaper access and tighter enterprise safeguards

Anthropic announced on September 1 that Claude Fable 5.1, one of its most capable generative AI models, is now generally available. The release comes with a revamped pricing structure that lowers costs for some research and high-volume users, alongside new privacy and safety features aimed squarely at enterprise customers. Fable 5.1 is a large language model designed for complex reasoning, long-context summarization and multilingual work, and Anthropic says it improves both capability and safety over the previous generation.

By contrast, Claude Mythos 5.1—an enhanced model for high‑risk areas such as cybersecurity and life sciences—will remain restricted to carefully vetted partners. The company is also debuting new “enterprise safeguards” that are meant to approximate zero‑data‑retention privacy while still allowing enough monitoring to detect abuse and improve the system over time.

The announcement underscores a broader shift in the AI industry: cutting‑edge models are no longer launched as pure capability upgrades. Vendors are expected to bundle pricing optimizations, governance controls and privacy guarantees, turning responsible deployment into a core part of the product story rather than an afterthought.