Claude Mythos 5.1 is Anthropic's most capable model, offered only to vetted cybersecurity and life sciences teams. It's the same model as Claude Fable 5.1, but with more permissive safeguards for that specialist work. For almost everyone else, Fable 5.1 is the version you can use, and it's generally available.
The key point: Mythos and Fable aren't two different brains. Anthropic says they're the same underlying model, with different levels of safeguards. So if you can't get Mythos, you're not missing out on a smarter model for everyday business work.
Everything here comes from Anthropic's own announcement of Claude Fable 5.1 and Claude Mythos 5.1, plus its Opus 5.5 announcement and help centre, which I read in October 2026.
Claude Mythos vs Claude Fable at a glance
| Claude Fable 5.1 | Claude Mythos 5.1 | |
|---|---|---|
| Underlying model | Same model | Same model |
| Who can use it | Generally available | Vetted organisations through trusted access programs |
| Safeguards | Standard, with fewer false positives than before | More permissive for cybersecurity and life sciences work |
| Where | All platforms, including AWS, Google Cloud and Microsoft Azure | A set of US organisations for now |
| Model name on the API | claude-fable-5-1 | Not open for general sign-up |
Who can get Claude Mythos
Anthropic says Mythos 5.1 will be available through two trusted access programs:
- The Cyber Verification Program (CVP). It already gives defensive security teams access to some Opus and Sonnet class models with reduced cyber safeguards, and Anthropic says it will soon include Mythos class models. Its Opus 5.5 announcement adds that the expanded program will have three tiers of increasingly permissive access.
- The Life Sciences Verification Program (LSVP). Built so life sciences professionals can use Mythos 5.1 for research and development, with every other safeguard still in place. Anthropic says it set this up in partnership with the US government and has enrolled its first participants.
Right now, Anthropic says Mythos is only available to a set of US organisations, while it works with the US government to widen access to more domestic and international partners. Claude Security, Anthropic's product that scans code for vulnerabilities and suggests patches for a human to review, is also powered by Mythos 5.1.
What you can do with Claude Fable 5.1 instead
Anthropic positions Fable 5.1 as its leading model for coding, knowledge work and long-running problem solving. Three details matter if you're thinking of using it:
- Default effort differs by app. Fable 5.1 defaults to High effort in Claude Code and Medium in Claude Cowork and on Claude.ai. Anthropic says at Low or Medium it matches or beats Fable 5 at a much lower cost.
- Cyber work is partly allowed. Fable 5.1 can now help discover software vulnerabilities, but not develop exploits for them.
- Biology research still routes elsewhere. Anthropic says research and development questions in the life sciences are still directed to its Opus models.
One practical catch: Anthropic's voice mode help article says Claude Fable isn't currently available in voice mode, so you'll need to type.
What Claude Fable 5.1 costs on the API
For developers paying by token, Anthropic lists Fable 5.1 at $10 per million input tokens and $50 per million output tokens, the same as Fable 5, in US dollars as of October 2026. The change is cache reads, where the model reuses context it's already processed: those now cost $0.25 per million tokens. Anthropic says that cut makes typical Fable workloads cheaper than Fable 5, with bigger savings on highly agentic work.
If you're on a Claude subscription rather than the API, my guide to how the Claude plans compare walks through your options, and what the Claude API is covers the developer side.
Privacy: Enterprise Frontier Safeguards
Alongside the launch, Anthropic introduced Enterprise Frontier Safeguards (EFS). Data is stored in cloud infrastructure the customer controls rather than on Anthropic's systems, giving the same privacy as a zero data retention policy while still detecting misuse. EFS is rolling out to enterprise customers in phases, and eligible customers can use Fable 5.1 with zero data retention until it's ready.
The watermark you can't see
Anthropic says it signed the EU AI Act's Code of Practice on Transparency of AI-Generated Content in July 2026, so models released after 2 August 2026, Fable 5.1 among them, add an invisible watermark to their text. Anthropic says it doesn't change the quality or content of what Claude writes, and only a detection tool can read it. That tool is in private preview for groups such as regulators, fact-checkers, researchers and educators.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
For business writing, nothing changes day to day. It's simply worth knowing that AI-written text can now carry a trace you can't see.
How to get access to Claude Mythos
- Check you fit one of the two programs: defensive cybersecurity work, or life sciences research and development.
- For cyberdefence, register interest through Anthropic's Cyber Verification Program.
- For biology research, apply to the Life Sciences Verification Program.
- Expect vetting, and expect access to start with US organisations.
If you're a small business owner, skip this. You're better served picking the right everyday model, and my guide to Claude Opus vs Sonnet walks through that choice.
My verdict
Claude Mythos is a specialist key for security teams and scientists, not a secret upgrade the rest of us are missing. Anthropic is clear it's the same model as Fable 5.1 with looser safeguards for vetted work. If you want the most capable Claude for your business, use Fable 5.1, set the effort level to suit the job, and save the hunt for Mythos access for when your work is in cybersecurity or the life sciences.
Frequently asked questions
Is Claude Mythos better than Claude Fable?
No, not as a model. Anthropic says Claude Mythos 5.1 and Claude Fable 5.1 are the same underlying model with different safeguards. Mythos has more permissive safeguards for vetted cybersecurity and life sciences work.
Can I use Claude Mythos?
Only if you're accepted into one of Anthropic's trusted access programs, the Cyber Verification Program or the Life Sciences Verification Program. For now, Anthropic says it's available to a set of US organisations.
Is Claude Fable 5.1 available to everyone?
Yes. Anthropic says Fable 5.1 is generally available on all platforms, including Amazon Web Services, Google Cloud and Microsoft Azure, and on the Claude API as claude-fable-5-1.
How much does Claude Fable 5.1 cost on the API?
Anthropic lists $10 per million input tokens and $50 per million output tokens, with cache reads at $0.25 per million tokens, in US dollars as of October 2026.
Does Claude Fable work in voice mode?
Not at the moment. Anthropic's help centre says Claude Fable isn't currently available in voice mode.
