Claude Fable 5.2 and Mythos 5.2 Launch With Major AI Upgrades

• Anthropic has introduced Claude Fable 5.2 and Claude Mythos 5.2 for demanding technical work.
• Fable 5.2 includes safeguards with fewer unnecessary blocks in early cybersecurity and biology tests.
• Mythos 5.2 is available to vetted organisations through Anthropic’s verification programmes.

Claude Fable 5.2 and Claude Mythos 5.2 are Anthropic’s latest AI models, designed for complex coding, research, cybersecurity and scientific work. The company says the two models share the same underlying model, but Fable includes additional safeguards for wider availability while Mythos is intended for vetted organisations working in sensitive fields.

The new models bring improvements in coding, reasoning and vision, along with faster output and lower pricing than their previous versions. Anthropic also reports fewer unnecessary safety blocks in early tests of Fable 5.2, while saying its cybersecurity and biology safeguards remain intact.

Claude Fable 5.2 Brings Stronger Coding and Research Capabilities

Anthropic says Claude Fable 5.2 is built for demanding, open-ended technical work. It has been used for software engineering, data-heavy research, hardware design, mathematics and robotics. The company says the model can handle ambiguous requests and coordinate multiple AI agents for large projects.

Anthropic also shared a specific figure about its own engineering work: more than 90% of the changes merged into Anthropic’s core codebase are now written by Claude, with human guidance. This refers to Anthropic’s own codebase and does not mean that the model writes 90% of all software code or works without human involvement.

In one internal project, a team of agents using an earlier version of Fable 5.2 migrated a production service from Python to Rust. Anthropic says the service became more efficient and reliable, and started around 26 times faster after the migration.

The company also describes an experiment in which Fable 5.2 coordinated more than 1,000 subagents over eight days to build a prototype operating system from an empty folder. The result included a working core, drivers, file storage, windows, a login screen and an installer. These are examples shared by Anthropic, rather than a guarantee of similar results for every coding task.

Claude Mythos 5.2 Targets Advanced Cybersecurity Work

Claude Mythos 5.2 is positioned as Anthropic’s most capable model for cybersecurity engineering. The company says it can identify more vulnerabilities than earlier models and improve the process of reproducing and fixing them. In Anthropic’s internal CVE Rediscovery evaluation, Mythos 5.2 found 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1.

Anthropic says a security company also found that one in six confirmed bugs identified by Mythos 5.2 had been missed by all the other frontier models included in its tests. These results come from specific evaluations and should not be treated as a guarantee that the model will find every security flaw.

Mythos 5.2 is not being opened to everyone. Cybersecurity professionals can apply through Anthropic’s Cyber Verification Program. Life sciences researchers can apply through its Life Sciences Verification Program. Anthropic Enterprise customers can also access Mythos 5.2 through Claude Security for codebase vulnerability scanning.

Fable 5.2 Reports Fewer Unnecessary Safety Blocks

Claude Fable 5.2 includes safeguards for high-risk cybersecurity activities and research biology tasks. Anthropic says these safeguards remain as robust as those in Fable 5.1, but its early tests found fewer unnecessary refusals or blocks when legitimate requests were made.

Compared with Fable 5.1, Anthropic observed around 60% fewer unnecessary blocks in cybersecurity and 40% fewer in biology. The company also reported about 55% fewer blocks overall across early testing on claude.ai, Claude Code and Cowork. These figures describe early test results, not a removal of safeguards or permission to carry out restricted tasks.

Fable 5.2 continues to block high-risk cyber requests, including penetration testing and exploit generation, as well as most research biology tasks. Mythos 5.2 is offered through the separate verification programmes because it is intended for organisations with legitimate specialist needs in these areas.

Claude Fable 5.2 and Mythos 5.2 Pricing and Availability

Anthropic lists standard pricing of $8 per million input tokens and $40 per million output tokens for Fable 5.2 and Mythos 5.2. An optional fast mode is also available at $16 per million input tokens and $80 per million output tokens. Anthropic says fast mode can increase token generation speed by up to 2.5 times.

The company says the new models generate outputs up to 30% faster than Fable 5.1 and Mythos 5.1, and are available at a lower price than those previous versions. Fast mode is available on Claude Code, the Claude Platform, Claude Models on Google Cloud and Microsoft Foundry. Mythos fast mode is limited to organisations in Anthropic’s verification programmes.

Claude Fable 5.2 is available across Anthropic’s platforms, including Amazon Web Services, Google Cloud and Microsoft Azure. Developers can access it through the Claude Platform using the model identifier claude-fable-5-2. Mythos 5.2 access remains limited to eligible organisations and enterprise security customers.

Feature	Description
Underlying model	Fable 5.2 and Mythos 5.2 share the same underlying model
Claude Fable 5.2	Includes cybersecurity and biology safeguards and is available across Anthropic’s platforms
Claude Mythos 5.2	For vetted cybersecurity and life sciences organisations, and eligible Claude Security customers
Standard API pricing	$8 per million input tokens and $40 per million output tokens
Fast mode pricing	$16 per million input tokens and $80 per million output tokens
Fast mode speed	Up to 2.5 times faster token generation, according to Anthropic
Model identifier	claude-fable-5-2 on the Claude Platform

Anthropic has published the model details, safety information and access conditions on its official Claude Fable 5.2 and Mythos 5.2 announcement page. Users and developers should check the company’s current platform documentation for account-specific access and pricing details.

MobileTelco has also covered other Anthropic updates, including the Claude legal artifacts gallery and the Claude Haiku 5.5 model and its lower-cost AI workloads.

Conclusion

Claude Fable 5.2 and Claude Mythos 5.2 extend Anthropic’s models for advanced coding, research and specialist technical work. Fable is the broadly available option with safeguards, while Mythos access is restricted to vetted organisations. The new pricing and early safety-test results are notable, but availability for Mythos depends on Anthropic’s approval programmes.

Frequently Asked Questions