Claude Fable’s Relaunch Leaves Users Underwhelmed by Subpar Performance
The long-awaited relaunch of Claude Fable, a powerful AI model from Anthropic, has been met with disappointment and frustration among users. Despite being made available to all users, including those with lower-tier subscriptions, the model’s performance falls short of expectations. According to reports, Fable 5 is severely capped on usage limits, and its capabilities are being restricted by strict safety systems.
When Claude Fable was first announced as returning from a ban imposed by the Department of Commerce, many were eager to see how it would perform. However, upon closer inspection, users have found that the model’s power has been significantly neutered. Those with higher-tier subscriptions can access Fable 5 for up to 50% of their weekly usage limits, but this cap will be even more restrictive after July 7, when the model transitions entirely to a pay-to-play system based on usage credits.
But it’s not just about the limitations – users are reporting that Fable 5 itself is performing below par. On Reddit, developers have been sharing their experiences with the model, which they describe as “weaker” and more prone to falling back to Opus 4.8 when faced with certain tasks or prompts. This shift in behavior has left many wondering if the original ban was simply a temporary measure, rather than a genuine safety concern.
The issue is not limited to the desktop version of Claude Fable; users have also reported similar problems with Claude Code, the cloud-based interface for accessing AI models. Some have even claimed that Fable’s “sensitivity” has increased, making it more likely to trigger fallbacks or block certain tasks altogether – especially when working with sensitive topics like security.
So what’s behind this underwhelming performance? According to Anthropic, the updated safety guardrails are to blame, rather than any changes to the model itself. These stricter safeguards aim to prevent potential misuse of AI models, but in doing so may inadvertently restrict their capabilities.
While it’s likely that Anthropic will address these issues in a future update, users are left to navigate this compromised version of Fable 5 for now. As with any powerful technology, it’s essential to remember that even the most advanced tools require careful handling and consideration of their limitations.
In light of this experience, one takeaway is clear: security teams should be prepared for the possibility that AI models may not always behave as expected. Regular testing and evaluation can help identify potential issues before they become major problems – a lesson that Anthropic would do well to take to heart in its ongoing efforts to refine its safety systems.
Source: Bleeping Computer — 2026-07-03