Claude Fable 5.2 is out. Try it on one hard job this week.

The story

Test it on one job while the launch offer runs

From November 9 through November 16, 2026 at 11:59 pm Pacific, Anthropic is including Fable models in the weekly limits on Pro plans and standard Team seats, for up to half of the week's limit, at no extra cost. After that, Fable on those plans goes back to usage credits, which are pay as you go. I'd use this week to pick one job that takes somebody hours and some judgment, run it on Fable 5.2 and on the model you use now, and compare the two. If what you have does the job, leave it alone.

Before you start

• Your default model doesn't change. Pick Fable 5.2 in the model picker.
• Fable models aren't on the Free plan.
• The launch offer ends November 16, 2026 at 11:59 pm Pacific.
• Fable uses your limits faster than other Claude models.
• Anthropic says to start with Opus 5.5 for most work.
• Have a person check its findings before anything leaves the office.

Have the person who does that job today judge the result, since they know what a right answer looks like.

How to try it yourself

You need a paid Claude plan and one finished job you already know the right answer to.

1. Pick Fable 5.2Open Claude and choose Fable 5.2 from the model picker. Your default model doesn't change, so you have to select it (Anthropic's help page (opens another site)). On a Team or Enterprise plan an owner controls which models are available, and on Enterprise plans Fable 5.2 starts out off, and an admin decides who gets it (Claude Fable models on your plan (opens another site)).
2. Give it a job you can gradeUse something finished, where you already know what's right: last month's statements against the books, or a contract you've read closely. Ask it what doesn't match and what's missing. Anthropic says early testers (opens another site) found it can handle more ambiguous, data-heavy tasks than previous models, and a job you've already done shows you whether that holds for your paperwork.
3. Run the same job on what you use nowGive the same files and the same question to the model you use every day. Anthropic calls Claude Opus 5.5 its strongest daily driver and recommends starting there for most coding and knowledge tasks. If both answers catch the same things, your everyday model is enough for that job.
4. Watch your usageSettings > Usage shows where you stand. During the launch offer, Pro plans and standard Team seats can use up to 50% of their weekly limit on Fable models. After you use that, or once the week ends, Fable usage is billed to usage credits if you have them turned on. Max and Ultra plans and premium seats include Fable in their weekly limits, up to the same 50% (Claude Fable models on your plan (opens another site)).
5. Mind what you uploadKnow where customer files go before you put them in. Anthropic's announcement says Fable relies on safeguards that monitor a window of usage data, and that for customers without a special enterprise arrangement it has 30-day retention for safety monitoring. The two questions I ask before AI sees customer data apply here too. For a first test, use your own business's paperwork.
6. Check its workHave a person check every figure and finding before it goes to a customer or a lender, because a better model still gets things wrong.
7. Leave alone what already worksIf a prompt or an automation already does its job on the model you have, keep it there. Move a job to Fable when your test shows it caught things the other model missed.