Get the app
LLMs

Fable 5.1 lands: same token price, much cheaper cache reads

Anthropic shipped Fable 5.1 to Claude and Claude Code. Per-token pricing is unchanged, but cheaper cache reads cut agentic bills up to ~45%.

Fable 5.1 lands: same token price, much cheaper cache reads

Anthropic pushed Fable 5.1 to Claude and Claude Code on 1 September, and the interesting part isn't the model card — it's the bill. Per-token pricing stays where Fable 5 was, but cache reads got substantially cheaper, which Anthropic puts at roughly 25% lower cost on typical token-billed workloads and up to ~45% on highly agentic work. The company also says 5.1 at low or medium effort matches or beats Fable 5 — so the cheap path got better, not just the expensive one.

On behavior, the pitch is reliability rather than benchmarks: writes in plain language and sticks to what you asked for, produces finished spreadsheets while checking each number as it goes, and separates what's known from what's estimated with sources shown. The flagship anecdote is a rare crash on Anthropic's internal systems that its own engineers — and every other model — had failed to explain after years.

Two catches. Fable 5.1 draws extra usage credits on Claude subscriptions, so the sticker price and the effective price aren't the same thing for plan users. And a restricted variant, Mythos 5.1, carrying different cybersecurity and life-sciences safeguards, ships only through Anthropic's trusted access programs.

Why it matters: when cache reads get cheap, long-running agents stop being a budget line item and start being the default.

Sources

Independent coverage

Written by an AI pipeline from the sources above. Methodology · Report an error

The daily AI brief, on your phone.

Feed, daily deep-dive and bytes — readable offline, with push alerts for the topics you follow.

Get it on Google Play