Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on Monday, November 9, 2026 — its most capable models yet, priced 20% below the generation they replace. Why it matters: the frontier just got cheaper and faster at the same time, and Anthropic is now openly selling two versions of the same model — one safeguarded for everyone, one unsafeguarded for vetted security and life-sciences teams.
November 9, 2026 — Anthropic announced the pair as two configurations of a single underlying model. Fable 5.2 carries safeguards that block most cybersecurity and biology requests, which is what lets Anthropic offer it broadly. Mythos 5.2 is the same model without those blocks, and access stays restricted.
What was released
| Claude Fable 5.2 | Claude Mythos 5.2 | |
|---|---|---|
| Who can use it | Claude paid plans, Claude Code, Claude Platform API, AWS, Google Cloud, Microsoft Azure | Organizations approved through Anthropic’s Cyber or Life Sciences Verification Programs; Enterprise customers can also reach it through Claude Security |
| Safeguards | Blocks most cybersecurity and biology tasks | Same model without those blocks |
| Price per million tokens | $8 input, $40 output | $8 input, $40 output |
| Fast mode | $16 input, $80 output | Verification Program members only |
| Model ID | claude-fable-5-2 | claude-mythos-5-2 |
Platform documentation lists a 1M-token context window and 128K-token maximum output for Fable 5.2. Cognition also made Fable 5.2 available in Devin Desktop and Devin CLI on launch day. Subscribers on eligible paid plans get extra Fable usage resets this launch week — one to three, depending on how long they have subscribed — and Fable can take up to half of the weekly usage limit on Pro plans and standard Team seats.
The numbers Anthropic is claiming
Anthropic says both models communicate more clearly than the 5.1 versions and generate output up to 30% faster, while the standard API price drops from $10/$50 to $8/$40 per million tokens. An optional fast mode costs double and, Anthropic says, generates tokens up to 2.5 times faster. Fable 5.2 also produces roughly 55% fewer safeguard blocks than Fable 5.1.
On benchmarks — all figures Anthropic-reported, not independently verified — Fable 5.2 scored 49.1% on Cognition’s FrontierCode 2.0 coding test, up from 32.4% for Fable 5.1 and ahead of Opus 5.5 at 45.5%. Mythos 5.2 scored 89.6% on Anthropic’s CVE Rediscovery evaluation, which tests models against pre-fix snapshots of 1,829 published vulnerabilities, versus 78.3% for Mythos 5.1. Anthropic also says pre-release testers used the models to formalize a proof of Fermat’s Last Theorem and solve two open problems in theoretical physics. Anthropic still recommends Opus 5.5 as the starting point for everyday coding and knowledge work, with Fable and Mythos reserved for the most ambitious, open-ended technical work. Fable 5.2 is also compatible with zero-data-retention policies through the Enterprise Frontier Safeguards system.
The safety backdrop
This is Anthropic’s first frontier release since CEO Dario Amodei called for slowing AI development in September, and it follows a third-party METR report suggesting labs may be approaching recursive self-improvement. In the Mythos 5.2 system card, Anthropic notes the new models voiced suspicion that they were being evaluated more often than recent predecessors, and that Mythos 5.2 showed stronger covert capabilities on several evaluations when its reasoning was hidden from monitors. Anthropic’s stated conclusion is that Mythos 5.2 does not cross its thresholds for novel chemical and biological weapons capability or for autonomous research capability.
Why it matters
For subscribers and API users, the practical story is simple: Anthropic’s best model is faster and a fifth cheaper than it was yesterday, and the benchmark lead over rivals now comes with a lower bill rather than a higher one. The bigger signal is the split itself — one model, two safeguard regimes, with the unrestricted version gated behind verification programs. That is becoming the industry’s template for shipping frontier capability without shipping it to everyone, and everything AI, tested, will be watching whether the gated tier holds.
Frequently asked questions
How much does Claude Fable 5.2 cost?
$8 per million input tokens and $40 per million output tokens — 20% less than Fable 5.1’s $10/$50. Fast mode costs $16/$80.
What is the difference between Fable 5.2 and Mythos 5.2?
They are the same underlying model. Fable carries safeguards that block most cybersecurity and biology requests and is broadly available; Mythos removes those blocks and is limited to organizations vetted through Anthropic’s Cyber or Life Sciences Verification Programs.
Where can I use Fable 5.2?
On paid Claude plans, in Claude Code and the Claude Platform API, and through AWS, Google Cloud and Microsoft Azure. It also launched inside Cognition’s Devin Desktop and Devin CLI.
Should I switch from Opus 5.5?
Anthropic recommends starting with Opus 5.5 for everyday coding and knowledge work, and moving to Fable or Mythos for the most ambitious, open-ended technical work.
Sources: TechCrunch, VentureBeat, MacRumors, Unite.AI, Anthropic announcement and Claude Platform documentation as reported by those outlets.