Many foreign media disclosed that Anthropic accidentally exposed a new model under testing due to draft content in the public cache. Public accounts show that Mythos has been described as one of the company's strongest models to date, while Capybara is more like a new hierarchical name higher than Opus; the final naming relationship between the two has not yet been officially announced.
Available information shows that the model has been tested with a small number of early customers, focusing on improving reasoning, programming and network security capabilities. The draft also emphasizes that its capabilities may pose significant cybersecurity risks, especially vulnerability exploitation and attack automation risks. The model has not yet been fully opened, and the release time, price, scope of application and risk control details are still unclear.
Frequently Asked Questions
Q: What model is Mythos?
A: Mythos is the name of the Anthropic new model under test disclosed by foreign media. The official naming has not yet been finalized.
Q: What level is Capybara?
A: The public draft shows that Capybara is more like a new model level higher than Opus.
Q: Who announced this news?
A: The news was first disclosed by foreign media, and Anthropic later confirmed to the media that limited testing was being carried out.
Q: Why does Anthropic emphasize risk?
A: Because this model has stronger network security-related capabilities, it may increase the risk of attack and abuse.
