Skip to content

NEWS · · 9 MIN

Claude Fable 5.2: price, benchmarks and comparison with Opus 5.5 and GPT-6 Astra

By AI GOATS Team

One task in the Artificial Analysis test cost Claude Fable 5.2 an average of 26.54 dollars. GPT-6 Astra did the same for 6.24 dollars, even though it charges more for every token. Yesterday Anthropic released the new model to all paying users, and today it already sits at the top of the independent ranking. We checked where the difference in the bill comes from, which tests Fable wins and how to set it up in Claude Code so you don't overpay.

In this article you will learn:

  • What changed in Fable 5.2 compared with version 5.1.
  • How much Fable 5.2 costs next to Opus 5.5, Sonnet 5.5, Haiku 5.5 and GPT-6 Astra.
  • How it performs in the Artificial Analysis ranking and in Anthropic's official tests.
  • Why a cheaper token does not mean a cheaper task.
  • Which effort level to choose in Claude Code and how to use the Fable week on the Pro plan.

Related articles

← All Academy resources