Claude Fable 5.1 is not included in the $20 Pro plan

Anthropic’s best model arrived on September 1, and the $20 plan gets a pay-per-use meter instead of access.

What launched

Claude Fable 5.1 and Claude Mythos 5.1 launched on September 1, 2026. Anthropic says Fable 5.1 scores 55.8% on Terminal-Bench 4.0 against 42.0% for Fable 5, and that cache reads now cost 75% less, which it estimates cuts typical workload costs by around 25% and agentic ones by up to 45%. Fable 5.1 is available on the API, in the Claude apps, and on AWS, Google Cloud and Azure. Mythos 5.1 goes only to vetted cyber and life-science programs.

The model is also doing unusual work. Vals AI handed it Sir Thomas Urquhart’s Cyphral Distich, a cipher from 1653 that had resisted solvers for 370 years and sits on Klaus Schmeh’s list of the top 50 unsolved encrypted messages. After 44 minutes it produced a reading: “O GOD UPHOLD KING CHARLS THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND.” Anthropic also used it to build a new elevation map of a third of Venus from 30-year-old Magellan radar data, at 2 to 3 km resolution instead of 10 to 20.

The plan fine print

Anthropic’s help center states that on Pro, Fable 5 and Fable 5.1 “aren’t included in your plan’s usage limits” and run on usage credits. On Max, subscribers can spend up to 50% of their weekly limit on Fable models.

Cost per token and cost per job also moved in different directions. Artificial Analysis measured the new model at $3.76 per Intelligence Index task, 20% more than Fable 5, because it writes about 1.7 times as many output tokens. The cache discount is real, but a longer answer eats part of it.

What it means for subscribers

The frontier moved again, and the $20 subscriber watches it from behind the rope. Anyone whose work lives on Pro now chooses between paying credits per use, upgrading to Max, or staying a version behind on purpose. The open question is whether pay-per-use access to the top model becomes the standard shape of consumer AI plans.