Urgent.News

What's breaking now, across thousands of outlets.

AI

Anthropic broadens Claude access with Fable 5.1

Anthropic has made Claude Fable 5.1 generally available while restricting the closely related Claude Mythos 5.1 to vetted partners, pairing stronger coding and research capabilities with new safeguards for cybersecurity, biology and enterprise data. The two models share the same underlying system, but Anthropic has separated access according to risk. Fable 5.1 carries tighter protections intended…

Anthropic has expanded Claude Fable 5.1's availability while limiting the more advanced Claude Mythos 5.1 to select partners. The two models share underlying system architecture, but Anthropic has separated access based on risk. Fable 5.1 is available to Pro, Max, Team, and Enterprise users through Anthropic's developer platform, cloud marketplaces, and major cloud providers.

It carries tighter protections for commercial use at a $10 per million input and $50 per million output token price. A significant pricing change is the reduced cache-read cost of $0.25 per million tokens, down from $1, which could lower costs by up to 45 per cent for certain workloads. Fable 5.1 boasts improved coding, research, and problem-solving capabilities while offering lower-cost performance.

It features a one-million-token context window and can generate outputs up to 128,000 tokens. Additionally, Anthropic has introduced Enterprise Frontier Safeguards (EFS), a system designed to prevent misuse while maintaining customer privacy. The company has also refined safety controls to reduce interventions in Claude Code sessions by 60 per cent.

For biology and life sciences, Anthropic has adjusted safeguards to intervene less often for benign queries while maintaining tighter controls for more advanced research. Mythos 5.1, intended for users needing fewer restrictions, is accessible through vetted programs for cyberdefenders and life scientists, with a 30-day data-retention period for safety monitoring.

Anthropic demonstrated the model's scientific reach through research tests, including the design of high-affinity protein binders and the acceleration of open-source biology models.

Written by urgent.news from Arabian Post's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at thearabianpost.com →

More in AI

More from Thursday 3 September →