Anthropic sparks strong user dissatisfaction over ClaudeCode performance degradation

On April 15, according to Fortune magazine in the United States, Anthropic was criticized by some core users for the performance decline of its Claude model. These issues have forced the company to cope with speculation about its motives and service capabilities as it reportedly prepares for an IPO. Many developers and heavy users say that Anthropic's Claude model has experienced significant performance declines recently, as evidenced by difficulty following instructions, choosing inappropriate shortcuts in complex tasks, and rising error rates. Most user complaints focus on Anthropic's programming tool ClaudeCode. The problem seems to be related to Anthropic's recent adjustments to how the model runs, which lowered the default "effort level" to reduce the number of tokens processed per request, thereby saving computing resources. There is speculation that Anthropic may face a shortage of computing power resources after the recent surge in its product use. Anthropic did not provide a detailed response to the complaint. Boris Cherny, product leader of ClaudeCode, said in an online response to user complaints that the company has adjusted the default "effort level" to "medium" in response to previous user feedback about excessive consumption of tokens, but some users believe that the company has not fully disclosed this change.