Skip to content
Anthropic

Claude Fable 5.2

Anthropic's top-tier model for the most ambitious, long-running technical work: cheaper and faster than Fable 5.1, and still a specialist next to Opus.

Quality

Modality

multimodal

Context

1M tokens

Access

closed

Fabian's Take

FM

"It's stronger than Opus, especially on really long-running tasks, but for almost all use cases Opus is probably the better choice. Several tests put Fable at roughly the same score as Opus at four times the cost, and Anthropic itself says to use Fable only for the most ambitious work."

Claude Fable 5.2 came out on 9 November 2026 and replaces Fable 5.1, which Anthropic now lists as a legacy model. Same 1M-token context window, same 128K output ceiling. Anthropic calls it a step up in coding, reasoning and vision.

What changed

The price came down: $8 per million input tokens and $40 per million output, from $10 and $50. Cache reads cost $0.20 per million, and batch requests are half price. Fast mode, which makes the output up to 2.5 times faster, costs $16 and $80 per million on the API, and you can switch it on in Claude Code.

By Anthropic’s numbers, 5.2 writes its output up to 30% faster than 5.1, and at medium effort, now the default in Claude Code and the Claude apps, it gets through hard tasks with fewer tokens. Code written against 5.1 needs some care: Anthropic lists four breaking changes.

Where it sits in the lineup

Anthropic is clear about the order: start with Claude Opus 5.5, “our strongest daily driver”, and switch only when you need to. Fable, in Anthropic’s words, is “best for the most ambitious, open-ended technical work”. At twice Opus’s price per token, that’s a narrow band of work.

My take: Fable 5.2 is stronger than Opus, especially on really long-running tasks. For almost everything else, Opus is probably the better choice. Several tests put Fable at roughly the same score as Opus at four times the cost.

The line has a history worth knowing about if you plan to depend on it. Fable 5 was pulled worldwide for nearly three weeks in June 2026 under a US government order before returning in July. I wrote about what that means for building on a single frontier model.

Where to find it

Fable 5.2 is in the Claude apps on paid plans, within usage limits that depend on the plan, in Claude Code, on Poe, and through the API, AWS, Google Cloud and Microsoft Azure. Same advice as since Fable 5: asking it simple things wastes money and usage. Route those to Opus or Sonnet.

Claude Mythos 5.2

Anthropic shipped a sibling, Mythos 5.2, on the same day. It’s the same underlying model with lighter safeguards for cybersecurity and biology work. Fable carries the stronger safeguards, which is why anyone can use it. Mythos is limited to vetted organizations in Anthropic’s Cyber and Life Sciences Verification Programs, so for almost everyone reading this it’s a name in an announcement rather than a choice to make.

The Verdict

Best for: The most ambitious, open-ended technical work and very long-running tasks, where Opus 5.5 falls short.

Pros

  • Stronger than Opus 5.5 on really long-running tasks
  • Cheaper than Fable 5.1: $8/$40 per million tokens, down from $10/$50
  • Output up to 30% faster than 5.1, and at medium effort it needs fewer tokens for hard tasks, by Anthropic's numbers
  • Same 1M-token context and 128K output ceiling as 5.1

Cons

  • Twice Opus 5.5's price per token, and Anthropic itself recommends starting with Opus
  • For almost every everyday job, Opus 5.5 is the better choice
  • Four breaking changes for anyone migrating code from Fable 5.1

Specs

  • Pricing $8/M input, $40/M output · cache read $0.20/M · batch 50% off · fast mode $16/$80
  • Cost Tier premium
  • ⚡
    Speed Tier slow
  • License Proprietary
Developer Docs

Access this model via