Anthropic, the artificial intelligence company behind the Claude family of models, has released Claude Fable 5.1 and Claude Mythos 5.1, its first update to the top tier of its model line-up since Fable 5 launched in June.
The two models are, according to Anthropic, the same underlying system with different safety layers attached: Fable 5.1 is generally available while Mythos 5.1 remains restricted to vetted organisations working in cybersecurity and life sciences.
The headline change is cost
Anthropic has cut pricing on cache reads, where the model processes previously stored input, from a dollar to 25 cents per million tokens, a 75% reduction.
That translates to roughly 25% lower costs on typical workloads and up to 45% on heavily agentic tasks, according to the company, while base input and output pricing stays unchanged at $10 and $50 per million tokens respectively.
On performance, Anthropic points to a sharp jump on Terminal-Bench-Science, a benchmark for agentic scientific research: Fable 5.1 scored 52.6%, against 24.7% for Fable 5, 29.0% for Opus 5 and 22.4% for OpenAI's GPT-5.6.
Anthropic itself cautions that the margin between models, rather than the ranking, should be treated carefully given a standard error of three and a half to four and a half points.
Coding and problem solving
The model is pitched at coding and long-running problem-solving work. Anthropic says Fable 5.1 avoids shortcuts that produce poor-quality fixes and is capable of tracking down the root causes of software issues.
As evidence, the company cites an internal test by the investment firm Millennium, in which Fable 5.1 identified the cause of a rare crash that had gone unexplained by the firm's own engineers, and every other model tested, for several years.
The release lands at a turbulent moment for Anthropic's flagship line. Fable 5 itself was pulled within days of its June launch under US export controls before being restored, and Anthropic's own Opus 5, released in July, had already undercut Fable 5 on price.
Cheaper and sharper?
Fable 5.1 is now available to Pro, Max, Team and Enterprise users, and to developers through the Claude Platform, Amazon Web Services, Google Cloud and Microsoft Foundry, priced at $10 per million input tokens and $50 per million output tokens.
For an industry increasingly judged on cost-per-token as much as raw capability, Anthropic is betting that a cheaper, sharper model will matter more than another headline benchmark.
