Anthropic has released two new AI models, Claude Fable 5.2 and Claude Mythos 5.2, which the company describes as its most capable large language models to date. The two models share the same underlying capabilities but differ in their safeguards: Fable 5.2 is broadly available, while Mythos 5.2 is restricted to vetted cybersecurity and life sciences organisations.

Both models are priced at $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for the previous generation, a cut of 20 per cent. Anthropic said the new versions also generate output up to 30 per cent faster than their predecessors.

Stronger results on coding and science

Anthropic reported gains across its tests. Fable 5.2 scored 49.1 per cent on the FrontierCode 2.0 benchmark at its highest effort setting, against 32.4 per cent for Fable 5.1, and 74.6 per cent on Terminal-Bench-Science 0.1, a 22 per cent improvement on its predecessor. In an internal experiment, an Anthropic engineer asked Fable 5.2 to build a prototype operating system; the model coordinated more than 1,000 subagents over eight days and produced software that could boot on a laptop.

Restricted access for Mythos 5.2

Mythos 5.2 is the version without Fable's biology and cybersecurity blocks, so Anthropic limits it to trusted researchers through its Life Sciences and Cyber Verification Programs. In the company's internal CVE Rediscovery evaluation, Mythos 5.2 identified 89.6 per cent of vulnerabilities, against 78.3 per cent for Mythos 5.1, across 1,829 previously disclosed flaws. Anthropic said customers eligible for zero data retention can keep those arrangements, while other customers face 30-day retention for safety monitoring. Read the Computerworld report and the SiliconANGLE report.