Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on November 9, positioning the models for difficult technical work. Standard API pricing is US$8 per million input tokens and US$40 per million output tokens; optional fast mode costs US$16 and US$80 respectively.
Tokens are the units of text a model processes or generates; the separate input and output rates make the shape of a workload important to its cost.
Shared model, different access
Anthropic says Fable and Mythos use the same underlying model, with Fable applying additional safeguards for cybersecurity and biology. Fable is available across its supported platforms. Organizations seeking Mythos can apply through verification programs; availability should not be read as unrestricted public access.
The company claims up to 30% faster output than the 5.1 versions, with optional fast mode increasing generation speed further. These are vendor-reported results, not independent measurements by Guava Insights.
Price per token is only part of the calculation
For a business, the useful unit is the cost of an accepted result. A cheaper token rate may not reduce a project’s bill if a workflow produces more output, makes more attempts or requires substantial review. Conversely, a more expensive run can be worthwhile if it reliably eliminates repeated work.
Evaluation should therefore hold the task and acceptance criteria constant, track unsuccessful attempts, and include human review time. The release expands the available options, but the appropriate model still depends on the workload and the permissions it needs. Faster generation alone does not demonstrate safer autonomous execution.
Sources
Original announcement · Developer migration and pricing documentation
Related coverage: Microsoft’s task-scoped permissions for AI agents.
Featured image: AI-generated conceptual illustration; not a depiction of an actual product or event.


