Anthropic released two new AI models on Monday: Claude Fable 5.2 and Claude Mythos 5.2. They are the company's most capable models so far for hard technical work — coding, research, and finding security flaws in software. Both are faster than the versions they replace, and both cost less.
There is a twist. Fable and Mythos are the same underlying model. The difference is who is allowed to use it, and how much freedom it has.
One model, two levels of access
Fable 5.2 is the public version. Anyone can use it through the Claude apps, the Claude API, Claude Code, and cloud platforms including Google Cloud, Microsoft Foundry and AWS. It keeps Anthropic's strongest safeguards against dangerous cybersecurity and biology requests.
Mythos 5.2 is the same model with fewer of those safeguards — because its job is security work itself. Finding real vulnerabilities in software means asking the questions the public version refuses to answer. So Mythos is not sold to the public. Access is limited to organizations approved through Anthropic's Cyber Verification Program or its Life Sciences Verification Program.
In short: one brain, two doors. Fable is the door everyone can walk through. Mythos is the locked door, opened only for vetted security and research teams.
The numbers that matter
Speed: both models generate output up to 30% faster than the 5.1 versions.
Price: $8 per million input tokens and $40 per million output tokens — a 20% cut from the $10 and $50 charged for the previous versions. An optional fast mode runs up to 2.5 times faster at $16 and $80.
Memory for work: a context window of 1 million tokens, with up to 128,000 tokens in a single answer — enough to read and work on very large documents and codebases.
Coding: Fable 5.2 scored 49.1% on Anthropic's FrontierCode 2.0 coding benchmark, up from 32.4% for Fable 5.1, and it ranked first overall on the independent Artificial Analysis Intelligence Index.
As a demonstration, Anthropic says a team of about 1,000 Fable 5.2 agents working together formalized the proof of Fermat's Last Theorem — a famous mathematics problem — in about 30 hours. An earlier version of the model needed 11 days for the same job.
Mythos, the bug hunter, is going into security products
The biggest practical news is where Mythos will show up. Most people will never touch the model directly — they will meet it inside security tools they already use.
Cisco announced that Mythos will power adaptive red teaming inside its AI Defense suite: it will plan attacks against a customer's own AI apps and agents to find weak spots before hackers do. Cisco customers get it under their existing licenses at no extra cost, and they cannot prompt the model directly — it works only on fixed tests against systems the customer owns.
IBM announced hours later that it will add Mythos to its Concert tool, which scans applications for risk. IBM says Mythos will "uncover weaknesses traditional scanners miss, determine which pose real risk and generate fixes for developer approval." IBM's version is planned for early 2027.
The numbers explain why security companies want it. On Anthropic's CVE Rediscovery test — finding known vulnerabilities in real software — Mythos 5.2 scored 89.6%, up from 78.3% for Mythos 5.1, and it found about 10% more vulnerabilities than Anthropic's Opus 5.5 model.
What this means for you
If you use AI tools to save time at work, Fable 5.2 is the version you will actually meet — in the Claude apps and in tools built on Claude. Expect quicker answers and, for developers paying per token, a lower bill.
If you build software, the price cut and the coding jump matter more than the headlines. Anthropic says Claude, with human guidance, now writes more than 90% of the changes merged into its own core codebase.
If you run security for a company, the strongest bug-finding AI is no longer something you apply for and wait on. It is arriving inside products from Cisco and IBM — with the model kept behind glass, so you get its findings without getting the model itself.
The bottom line
Anthropic's 5.2 release is really two stories in one. For everyone, Fable 5.2 is a faster, cheaper, stronger Claude. For a small group of vetted defenders, Mythos 5.2 is a specialist built to find the holes in software before attackers do — and it is now being packaged into the security tools of two of the oldest names in enterprise technology. The model is getting more powerful; the interesting change is how carefully Anthropic is choosing who gets to hold it.
Sources
Anthropic — Introducing Claude Fable 5.2 and Mythos 5.2, company announcement, 9 November 2026
Reuters — Anthropic unveils new Fable, Mythos models as competition heats up ahead of IPO, 9 November 2026
The Register — Anthropic's super bug-finding Mythos model is going into Cisco and IBM products, 10 November 2026
Digital Trends — Claude Fable 5.2 and Mythos 5.2 are here to raise the bar for AI coding and scientific research, November 2026
