Claude Fable relaunch disappoints users with nerfed performance
Anthropic relaunched Claude as "Claude Fable," but users report the new version has significantly reduced performance and capabilities compared to the previous model, leading to widespread disappointment in the AI community.
Background
- Claude is a family of AI models made by Anthropic, an AI safety company founded by ex-OpenAI employees. Its main rival is OpenAI's ChatGPT.
- "Fable" (likely a codename or internal project name) was a version of Claude that some users preferred for creative writing, storytelling, and role-playing tasks — domains where "uncensored" or less restricted models tend to perform better.
- "Nerfed" is slang for "nerfed" — meaning deliberately weakened or downgraded, often because of safety guardrails or performance trade-offs.
- Users are unhappy because the relaunched Fable model is reportedly less creative, more restricted, or worse at the tasks they previously relied on it for, which is a recurring tension in AI: safety/alignment vs. capability/expressiveness.
- This matters because power users (writers, developers, early adopters) often drive what features become mainstream, and their backlash can pressure companies into backtracking or explaining their decisions.