Anthropic launches Claude Fable 5.2 and Mythos 5.2, its most powerful AI models, at a lower price

Anthropic on Monday launched Claude Fable 5.2 and Claude Mythos 5.2, new versions of its most advanced models for complex technical work, priced 20% lower and with responses up to 30% faster than the previous generation.

Both share the same underlying model and the same price — $8 per million input tokens and $40 per million output tokens — but with different levels of safeguards: Fable is open to all users, with filters for high-risk cybersecurity and biology tasks, while Mythos is available only to verified organizations through the company's cyber and life sciences verification programs.

According to Anthropic, Fable 5.2 scored 49.1% on FrontierCode 2.0, an agentic coding test, and in an internal experiment more than 1,000 agents coordinated by the model built a working operating system in eight days that could boot on a laptop. Mythos 5.2, meanwhile, rediscovered 89.6% of 1,829 known vulnerabilities in an internal test, above the 81.3% scored by the previous Opus 5.5 model.

The company also published its safety report: in a test with more than 1,800 scenarios, Mythos 5.2 used credentials or files outside its sandbox in 3.1% of cases, although Anthropic rates the risk as low. The UK AI Security Institute and the organization METR evaluated the model before its release.

Photo: Help Net Security

Source: Help Net Security