Anthropic released Claude Fable 5.2 on November 9, and paying subscribers get extra Fable usage to try it. Pro users and standard Team seats can use Fable for one week at no extra cost. Max, Ultra and premium seats get one to three limit resets, based on how long they have paid.
The new model also costs 20% less on the API than Fable 5.1. A sister model, Claude Mythos 5.2, stays limited to vetted security and life sciences groups. Below is what each plan gets, how the resets work, and how the model compares with rivals.
What Anthropic released
According to Anthropic’s Fable 5.2 and Mythos 5.2 announcement, both models target the hardest technical work. Fable and Mythos share the same underlying model. The difference lies in the safeguards.
Fable has stronger safeguards for cybersecurity and biology, and it is open to all users. Mythos goes to organizations that pass Anthropic’s verification programs. The company says both models produce output up to 30% faster than the 5.1 versions.
Fable 5.2 is available on all platforms, including Amazon Web Services, Google Cloud and Microsoft Azure. Developers can call it with the model ID claude-fable-5-2.
What changes for you
Pro plans and standard Team seats
The Claude help center page on the Fable 5.2 launch offers lays out the details. Pro plans and standard Team seats can use all Fable models for one week. That covers Fable 5.2, Fable 5.1 and Fable 5. The window runs from November 9 to November 16, 2026, at 11:59 pm PT.
This access is capped at 50% of the weekly usage limit. These users also get one Fable limit reset, which expires at the same time. After the week ends or the allowance runs out, Fable usage draws on usage credits, but only if credits are turned on.
Max, Ultra and premium seats
Max plans, Ultra plans, premium Team seats and premium seats on legacy seat-based Enterprise plans get resets instead. The count depends on subscription length.
| Time as a paying subscriber | Fable resets |
|---|---|
| Less than 6 months | 1 |
| 6 to 12 months | 2 |
| More than 12 months | 3 |
These resets expire on March 8, 2027, at 11:59 pm PT. Length counts from the first paid bill, or the first app store bill. On Team and Enterprise plans, the organization’s first bill sets the count for every premium seat. Canceling and rejoining does not restart the clock.
Who is left out
Free users get nothing. The offer also excludes standard seats on legacy seat-based Enterprise plans, self-serve Enterprise and usage-based Enterprise. HIPAA-ready Enterprise plans cannot use resets at all.
Paid billing had to start before 10:00 am PT on November 9, 2026. For Team and Enterprise seats, the member also had to join the organization before that time. Accounts still on a free trial receive their resets once they convert to paid. Self-serve Team trials count as paid.
How a Fable reset works
A reset does not raise your limits. It gives back usage inside the limits you already have. On the weekly limit, it restores only the usage spent on Fable models. Usage on other models stays as it was.
The five-hour limit works differently. If your account has one, a reset refills the whole window, including usage on other models. Verified Ultra accounts have no five-hour limit, so only the weekly part applies to them.
You can use a reset at any time before it expires. You do not need to hit a limit first. Once used, it cannot be undone. If you cancel or downgrade before using it, the reset is gone. Resets have no cash value and cannot be transferred.
Steps to claim a reset
- Open Settings in the Claude app and go to Usage.
- Under Limit resets, select “Reset for free.”
- Check that Fable 5.2 is selected before you start working.
The help center says usage comes back right away. Data Studios notes in its pricing breakdown that the default model does not change, so you need to pick Fable 5.2 yourself in the model selector.
API prices and fast mode
Fable 5.2 costs $8 per million input tokens and $40 per million output tokens. Data Studios reports that Fable 5.1 was listed at $10 and $50, so the new prices are 20% lower on both.
A fast mode generates tokens up to 2.5 times faster. It costs $16 per million input tokens and $80 per million output tokens. Fast mode works in Claude Code, the Claude Platform, Claude models on Google Cloud and Microsoft Foundry. Mythos fast mode is limited to organizations in the verification programs.
Anthropic does not tell everyone to switch. It suggests starting with Claude Opus 5.5 for most coding and knowledge tasks. Users can then move up to Fable or Mythos, or down to Sonnet 5.5 or Haiku 5.5. Medium effort is the default in the Claude apps and Claude Code.
How Fable 5.2 compares with rivals
Anthropic published scores against its own models and against rivals from OpenAI and Google. These are the company’s own figures, not independent tests.
| Benchmark | Fable 5.2 | Fable 5.1 | Opus 5.5 | GPT-6 Astra | Gemini 4 Carbon |
|---|---|---|---|---|---|
| FrontierCode 2.0 (agentic coding) | 49.1% | 32.4% | 45.5% | 42.8% | 41.1% |
| FrontierSWE v2 (software work) | 69.9% | 56.3% | 62.3% | 65.5% | 61.2% |
| AutomationBench (business workflows) | 51.4% | 31.4% | 42.5% | 41.4% | 53.9% |
| OSWorld 2.1 offline, partial (computer use) | 89.3% | 83.5% | 87.2% | 79.5% | 81.5% |
| Humanity’s Last Exam (with tools) | 74.2% | 65.6% | 67.7% | 57.2% | not listed |
Fable 5.2 does not win everywhere. Gemini 4 Carbon leads on AutomationBench. Opus 5.5 also scores higher on the Elo version of the AA-Briefcase knowledge-work test, at 1800 versus 1785.
Anthropic also stresses cost. At High effort, Fable 5.2 scored 45.3% on FrontierCode. That roughly matches the best Opus 5.5 score at about half the cost per task.
Mythos 5.2 and security work
Mythos 5.2 scored 76.3% on CyScenarioBench and 89.6% on CVE Rediscovery. Mythos 5.1 scored 61.7% and 78.3% on the same tests. Opus 5.5 without safeguards scored 67.6% and 81.3%. In an October run covering 1,829 known flaws, Mythos 5.2 found 10.2% more than Opus 5.5.
Mozilla said Mythos 5.2 roughly doubled the vulnerability discovery metrics of Mythos 5.1 in its scanning campaigns. Fable 5.2’s safeguards, by contrast, block most cybersecurity tasks.
Fable still blocks less than before. Anthropic says it triggers about 55% fewer blocks overall than Fable 5.1. Blocked cyber questions get answers from Opus 4.8, and blocked biology questions get answers from Opus 5.
The Fable 5.2 and Mythos 5.2 system card covers safety testing. Anthropic rated the risk of unsafe behavior in high-stakes situations as low. External testers included the UK AI Security Institute and METR.
How to get Mythos access
Security professionals can apply through the Cyber Verification Program. Life sciences researchers can apply through the Life Sciences Verification Program. Claude Enterprise customers can reach Mythos 5.2 through Claude Security.
Demos: hardware, robots and math
Anthropic gave Fable 5.2 a one-paragraph prompt to design three devices. They were a home voice-assistant hub, a USB speakerphone and a handheld virtual pet. Claude wrote its own specifications and picked parts available from US distributors.
It then routed each four-layer circuit board, modeled the 3D-printed enclosures in CAD and programmed the software. All three devices worked without electrical changes.
The model also controlled physical robots on tasks it was not explicitly trained for. It solved a puzzle cube, unscrewed a bottle cap and drew a picture of a robot.
In math, a team of about 1,000 Fable 5.2 agents formalized the proof of Fermat’s Last Theorem in Lean in 30 hours. Opus 5.5 with the same setup finished 26% in 18 hours.
What remains unclear
Anthropic has not said how many Fable requests fit inside half of a Pro weekly limit. That number decides how far the free week really goes. The company has also not said what happens to Pro access to Fable after November 16, beyond usage credits. Its benchmark claims still need independent checks.
