Anthropic has launched Claude Fable 5.2 and Claude Mythos 5.2, its latest AI models for demanding technical work, describing them as faster and cheaper than the versions they replace. The company announced the pair on Monday, 9 November 2026, and said Fable 5.2 is available immediately across its platforms.

The two releases are built on the same underlying model. The difference is access, not capability. Fable 5.2 is the version for general use, with safeguards that limit work in cybersecurity and biology so it can be offered broadly. Mythos 5.2 is the same model without some of those restrictions, and it is available only to organisations approved through Anthropic's Cyber Verification and Life Sciences Verification Programmes. Anthropic announced the launch on its website, and Reuters reported the release on the same day.

Faster output and lower prices

Anthropic says the new models generate outputs up to 30 percent faster than Fable 5.1 and Mythos 5.1, and communicate more clearly with users. Standard pricing has fallen to 8 dollars per million input tokens and 40 dollars per million output tokens, down from 10 and 50 dollars for the previous generation.

An optional fast mode is available for 16 dollars per million input tokens and 80 dollars per million output tokens, which Anthropic says can speed up token generation by up to 2.5 times. Fable 5.2 has a one-million-token context window and supports up to 128,000 output tokens, allowing it to work with very large codebases and documents in a single session.

What the models are built for

Anthropic positions Fable and Mythos for the most ambitious, open-ended technical work, including long-running coding agents, scientific research, engineering and complex analysis. For most everyday coding and knowledge tasks, the company still recommends starting with Claude Opus 5.5.

In Anthropic's own published tests, Fable 5.2 scored 49.1 percent on FrontierCode 2.0 at its highest effort setting, the best result among the models the company tested. Mythos 5.2 reached 89.6 percent on Anthropic's internal CVE Rediscovery evaluation, which tests the rediscovery of known software vulnerabilities. These are the company's own figures, and they have not been independently verified.

Anthropic also reports striking results from pre-release testing, including that the models helped solve two previously open problems in black hole physics and designed three working electronic devices. It says a team of around 1,000 Fable 5.2 agents formalised the proof of Fermat's Last Theorem in 30 hours, a computer-checked formalisation rather than a new proof of the theorem itself. These claims come from Anthropic alone and remain to be checked by outside researchers.

Who can use Mythos 5.2

Access to Mythos 5.2 is deliberately limited. Organisations pursuing beneficial work in cybersecurity or biomedical research can apply through Anthropic's verification programmes, which the company says it has recently expanded. Enterprise customers can also use Mythos through Claude Security, which scans codebases for vulnerabilities.

Anthropic says Fable 5.2 keeps the cyber and biology safeguards of Fable 5.1 but produces fewer unnecessary blocks, and that it has added its strongest protections yet against distillation attacks, in which a rival model is trained on the outputs of another.

Fable 5.2 is available on the Claude Platform under the model ID claude-fable-5-2, as well as on Amazon Web Services, Google Cloud and Microsoft Azure. The launch comes as Anthropic prepares for a planned stock market listing, with Reuters reporting that the release is aimed at strengthening its position in the fast-growing AI market ahead of the IPO.