Test Claude Fable 5.2 on one hard job before you change your default.
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on Nov 9, 2026. Here is which one you can use, what it costs, how the launch offer works and a simple way to test it.
Approved Claude Partner SMB Trainer
What Anthropic announced
On Nov 9, 2026, Anthropic released two new models: Claude Fable 5.2 and Claude Mythos 5.2. Anthropic's announcement says that "Claude Fable and Claude Mythos are the same underlying model." The difference is who can use them. Fable has strong cybersecurity and biology safeguards to guard against misuse. Anthropic says those safeguards are what let it offer Fable to all users. Mythos 5.2 is only for vetted organizations that work in those dual-use areas and have been verified through Anthropic's verification programs.
Fable 5.2 or Mythos 5.2: which is which
- 01
Who can use it:
Fable 5.2 is open to all users on paid Claude plans, plus the API. Mythos 5.2 is only for organizations verified through Anthropic's verification programs.
- 02
Built for:
Fable 5.2 is for demanding reasoning and long agent work. Mythos 5.2 is for cybersecurity engineering; life sciences researchers can also apply.
- 03
API model ID:
claude-fable-5-2 for Fable 5.2 and claude-mythos-5-2 for Mythos 5.2.
- 04
API price per million tokens:
$8 input and $40 output for both models.
Anthropic calls Mythos 5.2 its most capable model for cybersecurity engineering. Cybersecurity professionals can apply for access through the Cyber Verification Program, and life sciences researchers through the Life Sciences Verification Program. Most small businesses will not be in either group, so Fable 5.2 is the model you can use.
What Anthropic says it is for
Anthropic's Fable 5.2 overview says the model is built for "demanding reasoning and long-horizon agentic work." The overview lists typical uses: agent sessions that run for hours, multistep deep research, and analysis carried through to a finished document, spreadsheet or deck.
Anthropic's announcement includes reports from its testers at larger companies. Here are some of the business examples:
- 01
I-9 compliance review.
Niko Grupen, Head of AI Research at Harvey: "Claude Fable 5.2 gave the client clearer direction than Fable 5.1."
- 02
Loan application files.
Yashodha Bhavnani, VP of AI Products at Box: "it flagged that the salary deposits and the bank statement's balance didn't match."
- 03
10-K financials into a forecasting model.
Jeffrey Liu, Member of Technical Staff at Fundamental Research Labs: it "caught that the regional numbers didn't sum to the reported total, and got every cell right."
- 04
Enterprise SQL.
Liyu Gong, Senior Manager of Machine Learning at Oracle: "Claude Fable 5.2 generated SQL that returned the expected results."
These testers were reporting on their own work. None of them tested the model on a small business, and their results don't tell you how Fable 5.2 will handle your files.
How it compares
Every figure below comes from Anthropic's benchmark table in its announcement. These numbers come from Anthropic, and Consultiply has not re-run any of them. The table's footnote says scores for Fable 5.2, Fable 5.1 and Opus 5.5 "show results with safeguards on unless otherwise noted. All scores show each model's highest-available effort unless otherwise noted."
- 01
FrontierCode 2.0 (agentic coding):
Fable 5.2 scored 49.1% at Xhigh effort. Opus 5.5 scored 45.5% at its highest available effort. Anthropic adds that at High effort, Fable 5.2 scored 45.3%, "for roughly half the cost per task."
- 02
Remote Labor Index (knowledge work):
Fable 5.2 scored 41.3% at High effort. Opus 5.5 scored 21.3%.
Anthropic sums up the table by saying the two new models "are a step above Opus 5.5 in almost all the capabilities we evaluated." The same table shows two exceptions, also at highest available effort. On AutomationBench (business workflows), Gemini 4 Carbon scored 53.9% and Fable 5.2 scored 51.4%. On the AA-Briefcase v1.1 Elo row, Opus 5.5 scored 1800 and Fable 5.2 scored 1785.
For most coding and knowledge tasks, Anthropic still recommends that users "start with Claude Opus 5.5, our strongest daily driver," and switch models when the work calls for it.
How to get it
The Claude Help Center says Fable models are on all paid plans: Pro, Max, Ultra, Team and Enterprise. Fable is not available on the Free plan. How you get it depends on your plan or seat:
- 01
Max, Ultra and premium Team seats
(plus premium seats on legacy seat-based Enterprise plans): Fable is included for up to 50% of your weekly usage limits at no extra cost. It uses up those limits faster than other Claude models.
- 02
Pro and standard Team seats:
Fable isn't part of your usage limits. You pay for it with usage credits, except during the launch week covered below.
- 03
Enterprise:
Fable 5.2 stays off until an admin turns it on. Seven days after launch, Anthropic adds it to every role that already has a Fable model turned on, unless an admin opts out first.
In Claude on the web, Desktop and Mobile, you choose Fable from the model picker. The Help Center says "Your default model doesn't change." Claude Code needs version 2.1.321 or later.
Developers can call claude-fable-5-2 on the Claude API. Anthropic's docs also list it on Amazon Bedrock, Claude Models on Google Cloud, Microsoft Foundry and Claude Platform on AWS. The announcement says "Microsoft Azure" where the docs say Microsoft Foundry.
The Anthropic pages we used don't give monthly dollar prices for Pro, Max, Ultra or Team, or a price for usage credits.
What it costs
Anthropic's pricing page lists the same API prices for Fable 5.2 and Mythos 5.2, in USD per million tokens. Fable 5.1 and Mythos 5.1 are shown for comparison:
- 01
Input:
$8 for Fable 5.2 and Mythos 5.2, against $10 for the 5.1 models.
- 02
Output:
$40, against $50.
- 03
Batch input and output:
$4 and $20, against $5 and $25.
- 04
Cache read:
$0.20, against $0.25.
- 05
Fast mode input and output:
$16 and $80. Not stated for the 5.1 models.
We did the math on those printed prices: Fable 5.2 is 20% cheaper than Fable 5.1 on every line where both are listed. Fast mode costs twice the standard rate, and the Batch API doesn't support it.
The launch offer, Nov 9 to Nov 16
These are the terms on the Claude Help Center launch offer page:
- 01
Pro plans and standard Team seats:
all Fable models are included from Nov 9, 2026 through Nov 16, 2026 at 11:59 pm PT, for up to 50% of weekly usage limits. You also get one Fable limit reset, which expires Nov 16, 2026 at 11:59 pm PT.
- 02
Max, Ultra and premium seats
(premium Team seats and premium seats on legacy seat-based Enterprise plans): one to three Fable limit resets, depending on how long since your first paid bill. Under six months gets one reset, six to 12 months gets two, and more than 12 months gets three. Team and Enterprise plans are counted at the organization level. These resets expire Mar 8, 2027 at 11:59 pm PT.
- 03
Eligibility:
paid billing had to be in place before Nov 9, 2026 at 10:00 am PT, and seat holders must have joined before then. The offer doesn't cover the Free plan, standard seats on legacy seat-based Enterprise plans, or self-serve, usage-based or HIPAA-ready Enterprise plans. Resets have no cash value and can't be transferred.
- 04
What a reset does:
it restores your Fable usage on the weekly limit and leaves your usage of other models as it is. If your account has a five-hour window, the reset refills all of it, other models included. A reset does not raise your limits.
Try it this week
These steps are our suggestions, not Anthropic's. They don't claim anything about what the model will do for you.
- 01
Step 1.
Pick one hard, document-heavy job where you already know what a good answer looks like.
- 02
Step 2.
Run it on your current model first and save the output.
- 03
Step 3.
Run the same job on Fable 5.2 at the default effort for where you're using it. Anthropic's docs list high as the API default. Its announcement says medium is the default in the Claude apps and Claude Code.
- 04
Step 4.
Put the two outputs side by side and check every figure yourself. Then decide who in your business can turn Fable on and what data it can see. Anthropic's docs say both models "require 30-day data retention and aren't available under zero data retention unless expressly authorized by Anthropic." If a managed IT provider looks after your accounts, involve them in that decision.
Our take
Before Fable 5.2 becomes part of your routine, test it once, side by side, on a real job. For everyday work, Anthropic's own advice to start with Opus 5.5 makes sense, and Fable is worth a try on long, hard jobs whose results you can check yourself.
A Claude workshop in Tulsa
Consultiply, an Approved Claude Partner SMB Trainer, is hosting a Claude SMB Workshop with TCC Skills Lab in Tulsa on Mon Dec 14, 2026. Apply for a seat.
Sources
Claude Fable 5.2 and Claude Mythos 5.2 (Anthropic, announcement, Nov 9, 2026)
What's new in Claude Fable 5.2 (Anthropic, Claude Docs)
Claude Fable 5.2 overview (Anthropic, Claude Docs)
Claude Mythos 5.2 overview (Anthropic, Claude Docs)
Pricing (Anthropic, Claude Docs)
Claude Fable 5.2 launch offers (Claude Help Center)
Claude Fable models on your plan (Claude Help Center)
Want this working in your business?
Tell us what you are trying to fix and we will point you at the smallest useful next step.
Talk to us