All three of these are big projects that would normally take anywhere from hours to days or months. Instead, each one was made with a one-shot prompt to Fable 5, the new model out today from Anthropic. And it’s not just vibe-coded prototypes: Fable 5 scored a 91/100 on our Senior Engineer benchmark compared to Opus 4.8’s 63 and GPT-5.5’s 62. Anthropic describes it as the most capable model it has ever made available to the public, and we agree. (View Highlight)
It is also the first of what the company is calling Mythos-class models. You’ve undoubtedly heard the word Mythos before—Anthropic declared Mythos too dangerous for public use in April and restricted access to a handful of partners. Since then, the company has given limited release to a chosen set of large companies for feedback. The model released today is a version with safeguards that prevent it from being used for anything related to cybersecurity or biology. (View Highlight)
In our testing, we found that users who were highly adept with AI—at Level 7 or 8 on our AI adoption ladder—found it paradigm-shifting for their hardest tasks. Users who were lower down on the curve, however, struggled to find something to use it for. (View Highlight)
Seven of us have been testing it for about a week at Every—putting it through its paces for coding, writing, business strategy, data analysis, and growth. This is our day-zero Vibe Check. (View Highlight)
The general-release model comes with unusual guardrails. Anthropic says Fable will block some requests involving cybersecurity and biology (bad news for Dan and his health-tracking habits); most of those queries will be routed to Claude Opus 4.8 instead. Anthropic is also releasing Mythos 5, the same base model without the cyber and biology safeguards, but access will remain limited. For most users, Fable is the version they will encounter. (View Highlight)
