AI Generated Image
Claude Fable relaunch disappoints users with nerfed performance
BleepingComputer • • business
Key Points:
- Claude Fable, Anthropic's most powerful AI model, is now available to all users but with significant usage restrictions, including a 50% weekly usage cap for Max, Pro, and Team plans, and a planned shift to a fully pay-to-play system after July 7.
- Users report that the restored Fable 5 model shows degraded or "nerfed" performance, frequently triggering fallback to the less capable Opus 4.8 model due to stricter safety guardrails.
- The stricter safety measures cause the model to reject or downgrade tasks related to security, system-level coding, and certain sensitive keywords, making it less effective for some technical uses.
- Despite user complaints, the core model itself has not been weakened; rather, Anthropic's enhanced safety protocols with a large "safety margin" are likely causing false positives and reduced usability.
- Anthropic has not yet publicly responded to reports of these issues but is expected to address the false positive and performance concerns in future updates.
