Anthropic releases Claude Fable 5.1, and makes it cheaper to run
The new frontier model lands with a large cut to cache-read pricing, and a restricted sibling aimed at cybersecurity and life-sciences work.
Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on 1 September 2026, describing them as its most capable models for coding and knowledge work. The two are the same underlying model with different safeguards: Fable 5.1 is generally available, while Mythos 5.1 is restricted to trusted access programmes, with protections aimed at cybersecurity and life-sciences work.
The headline for anyone paying by the token is cost rather than capability. Anthropic estimates Fable 5.1 will cost roughly 25% less than Fable 5 for typical workloads, and VentureBeat reported a 75% reduction in the price of cache reads — the part of the bill that dominates long agentic sessions, where the same context is read over and over.
It is available as claude-fable-5-1 through the Claude API, Amazon Bedrock, Google Cloud and Microsoft Foundry. MarkTechPost reported a score of 52.6% on Terminal-Bench-Science alongside the pricing changes.
Like every frontier model from the major labs, Fable 5.1 is served from the provider’s own infrastructure. There are no weights to download and nothing to run on your own machine, so it does not appear in this site’s catalogue — which covers open-weights models you can run yourself.
Sources
This is a summary of reporting published elsewhere. Where a claim matters, follow the source — it is the record, this is not.
Which models can your PC run?
Scan your hardware and see every model that fits, with the memory each one needs.