Anthropic leak reveals new model "Claude Mythos" with "dramatically higher scores on tests" than any previous model
Update –
- Added more details on the leaked model documents.
Update March 27, 2026:
The leaked blog posts have allegedly surfaced online; the information matches what Fortune shared in a follow-up article. There are two versions of the same blog post that only differ in the model's name: "Mythos" (v1) and "Capybara" (v2). Anthropic told Fortune the documents were simply "early drafts of content that were being considered for publication," suggesting the company was deciding between two name candidates for the same model.
In the Capybara version, the name was swapped throughout the title and body text, but the subtitle still reads, "We have finished training a new AI model: Claude Mythos." Both versions also use the same justification for the name, saying it was chosen to evoke "the deep connective tissue that links together knowledge and ideas."
According to the drafts, the model represents a new class above Anthropic's existing Opus line, described as "larger and more intelligent than our Opus models—which were, until now, our most powerful," achieving "dramatically higher scores on tests of software coding, academic reasoning, and cybersecurity, among others" compared to Claude Opus 4.6.
When asked for comment, Anthropic confirmed to Fortune that it's training and testing a new model. "We're developing a general purpose model with meaningful advances in reasoning, coding, and cybersecurity," a spokesperson said. "We consider this model a step change and the most capable we've built to date."
Anthropic plans a slow, security-focused rollout
According to the documents, Anthropic is planning a deliberately slower release than with previous models. The model is allegedly "currently far ahead of any other AI model in cyber capabilities," but "presages an upcoming wave of models that can exploit vulnerabilities in ways that far outpace the efforts of defenders." The rollout will start with a small group of early-access customers tasked with evaluating cybersecurity applications, with access through the Claude API expanding gradually.
The drafts also acknowledge the model is "very expensive for us to serve, and will be very expensive for our customers to use." Anthropic says it's working to make it "much more efficient before any general release." Whether the model ultimately ships as "Mythos," "Capybara," or under a different name remains to be seen.
Original short report:
A data leak at Anthropic has exposed details about an unreleased AI model that internal documents call the company's most powerful to date. After Fortune broke the story, Anthropic confirmed it is already testing the model with select customers, claiming it marks a "step change" in reasoning, coding, and cybersecurity capabilities. The breach happened because of a misconfiguration in Anthropic's content management system. A default setting automatically made uploaded files public, leaving nearly 3,000 internal documents exposed for anyone to see.
OpenAI is reportedly also gearing up for a major release. The company is preparing a new model codenamed "Spud," which has already finished pretraining. Similar to Anthropic, OpenAI CEO Sam Altman has internally promised a massive jump in capabilities, saying the model can "really accelerate the economy," whatever that means. Both companies will likely time the release of their strongest models to ensure they are optimally positioned for their planned IPOs later this year.
AI News Without the Hype – Curated by Humans
As a THE DECODER subscriber, you get ad-free reading, our weekly AI newsletter, the exclusive "AI Radar" Frontier Report 6× per year, access to comments, and our complete archive.
Subscribe nowAI news without the hype
Curated by humans.
- Over 20 percent launch discount.
- Read without distractions – no Google ads.
- Access to comments and community discussions.
- Weekly AI newsletter.
- 6 times a year: “AI Radar” – deep dives on key AI topics.
- Up to 25 % off on KI Pro online events.
- Access to our full ten-year archive.
- Get the latest AI news from The Decoder.