Claude Fable 5.2 and Mythos 5.2 arrive

Anthropic announced Claude Fable 5.2 and Claude Mythos 5.2 on November 9, 2026. The company says both models are designed for demanding technical work and bring improvements in coding, reasoning and vision. Standard pricing is $8 per million input tokens and $40 per million output tokens.

According to @claudeai, the models are built to tackle ambitious projects across many fields and are meant to be capable collaborators. In addition, Anthropic says Fable 5.2 is available to everyone, while Mythos 5.2 is limited to verified organizations.

What is the difference in Claude Fable 5.2 and Mythos 5.2?

Anthropic says both models share the same base model. However, their access policies differ. For example, Fable 5.2 has stronger safeguards against cybersecurity and biology misuse and is publicly available. Mythos 5.2 is offered only to verified organizations through the Life Sciences Verification Program and the Cyber Verification Program.

The company also says output generation is up to 30% faster than Fable and Mythos 5.1. As a result, an optional fast mode can speed up token generation by up to 2.5x, priced at $16 per million input tokens and $80 per million output tokens.

How much does Claude Fable 5.2 cost and where is it available?

Anthropic recommends starting with Claude Opus 5.5 for most daily coding and knowledge work, then moving up to Fable or Mythos as the task gets larger. Meanwhile, Sonnet 5.5 and Haiku 5.5 remain better for simpler jobs. Fable 5.2 can also be used in Advisor or Orchestrator roles. The API ID is claude-fable-5-2, and the model is available today across AWS, Google Cloud and Microsoft Azure.

Turkey pricing and local availability were not announced. In addition, Anthropic said paid plans will get extra Fable usage quota this week as part of a promotion.

What Claude Fable 5.2 performance claims did Anthropic publish?

In its own tables, Anthropic says Fable 5.2 scored 49.1% on FrontierCode 2.0 Xhigh, 69.9% on FrontierSWE v2, and 74.2% on the tool-using Humanity’s Last Exam benchmark. For Mythos 5.2, the company published 75.3% on Terminal-Bench-Science 0.1 and 76.3% on CyScenarioBench. However, these are Anthropic’s own tests, and this article does not include independent third-party verification.

On safety, Anthropic says external evaluations with the UK AI Security Institute, Frontier Design and METR showed about 55% fewer false positives in Fable 5.2. For example, the company says the reduction is about 60% in cyber and about 40% in biology. Its Enterprise Frontier Safeguards private beta includes Bridgewater, Salesforce and Stripe.

Model	Price	Access	Note
Claude Fable 5.2	$8 / $40	Public	claude-fable-5-2
Claude Mythos 5.2	$8 / $40	Verified organizations	Verification programs
Fast mode	$16 / $80	Optional	Up to 2.5x faster token generation

Read our previous coverage of Claude Fable 5.1, our story on GPT-6 Cyber and our piece on Claude personal finance. Anthropic published the official announcement on November 9, 2026, and @claudeai also posted a launch video on X.