[AIbase Report] Advanced Users of Claude Code Encounter Unannounced Usage Restrictions, Anthropic Provides Vague Response
Since this week, the Claude Code service from Anthropic has suddenly faced usage restrictions, causing dissatisfaction among heavy users. Especially the Max subscription users who pay $200 per month have been frequently receiving "Claude usage limit reached" notifications without any prior notice, leading to interrupted project progress and inability to use the service normally.
Users have concentratedly reflected these issues on the GitHub page of Claude Code and questioned whether Anthropic has changed the usage tracking mechanism. One user said, "In the past 30 minutes, I didn't send 900 messages at all, but I have already been restricted." The Max plan was originally promised to offer 20 times the resources of Pro users, while the Pro plan offers five times that of the free version. Despite this, Anthropic has not clearly guaranteed any specific usage limits for each subscription tier, making it difficult for users to plan their work stably.
Anthropic has not provided a substantive explanation, and only one representative responded by saying, "We have noticed some Claude Code users experiencing slower response speeds and are working on fixing it." However, the company's status page shows six technical failures in the past four days. Although the stated uptime is still 100%, users report frequent overload errors.
An anonymous user told TechCrunch that due to the restrictions on Claude Code, his entire project has come to a standstill. He tried using alternatives like Gemini and Kimi, but admitted, "No tool can replace Claude Code at the code level."
Additionally, users expressed disappointment with Anthropic's opaque pricing strategy. "As long as they are transparent, it would be fine," said a Max user. "The current approach will only make people lose confidence in them."
As the issues of service restrictions and poor communication expand, the future operational strategy of Claude Code and user trust are facing significant challenges.