Excessive token usage in Claude Code (github.com)

Users of the Claude Code AI platform have reported a significant increase in token usage following the recent update to version 2.1.1. Many users, who previously experienced normal usage levels, are now hitting their token limits multiple times faster than before, leading to frustrations about the efficiency of the system. One user noted that their token consumption has skyrocketed to levels exceeding double the holiday season promotion rates, with usage spiking dramatically for simple tasks that previously required minimal resources. This issue is particularly significant for the AI/ML community as it highlights potential inefficiencies and scalability concerns within the Claude Code platform. The rapid consumption of tokens could hinder developers' workflows and project productivity, especially for those on tight budgets or usage caps. Technical details indicate that errors such as "Lock acquisition failed" and aborted requests are also contributing to the frustrations, suggesting that the update may have introduced underlying stability or resource management issues that need to be addressed promptly. Addressing these concerns will be crucial for maintaining user trust and ensuring continued adoption of AI tools in development practices.