Claude Mythos 5.2
- Provider
- Anthropic
- Status
- Restricted
- Price
- $8 / $40 /MTok
Claude Mythos 5.2 is Anthropic’s restricted Mythos-class model, released on 9 November 2026 alongside Claude Fable 5.2 (Anthropic). Anthropic says the two are the same underlying model: Mythos 5.2 is Fable 5.2 without certain cybersecurity and biology safeguards. It succeeds Claude Mythos 5.1.
Access is limited to vetted organisations working in cybersecurity or biomedical research. Anthropic’s pricing page lists it at $8/$40 per million tokens, the same as Fable 5.2, marked limited availability. There is no independent score: Artificial Analysis scored Fable 5.2 at 65 on its Intelligence Index v5.0.1 at launch, but does not list Mythos 5.2.
Quick specs
| Provider | Anthropic |
| Released | 9 November 2026 (vetted organisations only) |
| Same model as | Claude Fable 5.2, with fewer cyber and biology safeguards |
| Access | Cyber Verification Program, Life Sciences Verification Program, or Claude Security on Claude Enterprise |
| Price | $8 / $40 per MTok (limited availability) |
| Fast mode | $16 / $80 per MTok, verification programme members only |
| Independent score | None |
| SWE-bench | Not published |
Who can use it
- Cybersecurity professionals can apply through the Cyber Verification Program.
- Life-science researchers can apply through the Life Sciences Verification Program.
- Claude Enterprise customers can use Mythos 5.2 inside Claude Security without joining the Cyber Verification Program.
Everyone else gets Fable 5.2, where blocked cyber requests are answered by Claude Opus 4.8 and blocked biology requests by Claude Opus 5.
Benchmark performance
| Benchmark | Mythos 5.2 | Comparison |
|---|---|---|
| Terminal-Bench-Science 0.1 | 75.3% | Fable 5.2 74.6%, Opus 5.5 58.7% |
| CVE Rediscovery | 89.6% | Mythos 5.1 78.3%; Mythos 5.2 with safeguards 81.3% |
| CyScenarioBench | 76.3% | Mythos 5.1 61.7% |
| De novo protein binder design (24 hours) | 85.1% | Mythos 5.1 79.1%; Mythos 5.2 with safeguards 82.6% |
All figures are Anthropic’s own. Anthropic says that in an October 2026 run over 1,829 known vulnerabilities, Mythos 5.2 found 10.2% more than Opus 5.5 and 14.4% more than Mythos 5.1.
Safety results
Anthropic rates the risk of unsafe behaviour in high-stakes situations as low. In its automated behavioural audit of about 1,900 scenarios, Mythos 5.2 scored 2.67 (lower is better), against 2.79 for Opus 5.5 and 3.28 for Mythos 5.1. In more than 1,800 adversarial sandbox scenarios it used outside credentials or files in 3.1% of cases, against 21% for Mythos 5.1. The UK AI Security Institute and METR were among the external evaluators (Anthropic).
Known limitations
No independent measurement. Artificial Analysis does not list Mythos 5.2, so every figure for it is Anthropic’s.
Gated access. No public API and no consumer access outside the verification programmes and Claude Security.
Missing specs. Anthropic’s models overview does not list Mythos 5.2, so its API model ID, context window and knowledge cutoff are not published.
Frequently asked questions
When was Claude Mythos 5.2 released?
9 November 2026, alongside Claude Fable 5.2.
How is Mythos 5.2 different from Fable 5.2?
Anthropic says they are the same underlying model. Mythos 5.2 lacks certain cybersecurity and biology safeguards that Fable 5.2 has, which is why it is limited to vetted organisations.
How do I get access to Claude Mythos 5.2?
Apply through Anthropic’s Cyber Verification Program or Life Sciences Verification Program, or use it through Claude Security on a Claude Enterprise plan.
How much does Claude Mythos 5.2 cost?
Anthropic’s pricing page lists it at $8 per million input tokens and $40 per million output tokens, the same as Fable 5.2, for admitted organisations. Fast mode costs $16/$80 and is only for verification programme members.