Anthropic Launches Claude Fable 5.2 and Mythos 5.2 Models

New Models

Anthropic released two new models on Monday, November 09, named Claude Fable 5.2 and Claude Mythos 5.2. This is the first release of frontier models by the company since its call for a slowdown in development. The rollout also happens as the company prepares for a historic initial public offering that is expected to happen sooner.

Performance And Pricing

Anthropic states that both models are its most advanced options yet. The new offerings run up to 30 per cent faster than their older versions and cost about 20 per cent less than Fable and Mythos 5.1. Both models carry a price of $8 per million input tokens and $40 per million output tokens. They show better results in coding, reasoning, and vision. During testing, the models helped engineers complete difficult software migrations, design three working electronic devices from scratch, solve two open problems in black hole physics, make progress on robotics evaluations, and formalise the proof of Fermat's Last Theorem in 30 hours.

Differences And Availability

Fable 5.2 and Mythos 5.2 share the same underlying model but differ in safety rules and who can use them. Fable 5.2 includes safeguards that restrict potentially dangerous cybersecurity and biological research uses, which allows it to reach a broader audience. Mythos 5.2 has fewer limits and is meant only for vetted groups working on cybersecurity and biomedical research. Fable 5.2 is available now across the company platforms like Claude, Claude Code, and its API, as well as through Amazon Web Services, Google Cloud, and Microsoft Azure. Mythos 5.2 is not generally available yet.

Context

Anthropic is one of the leading artificial intelligence research companies that develops large language models and software tools. Companies in this sector regularly release updated versions of their models to offer faster speeds, lower costs, and better reasoning capabilities for developers and enterprise users.