[Anthropic Faces Strong User Dissatisfaction Over Claude Code Performance Decline] Jin10 Data, April 15 – According to a report by the U.S. magazine *Fortune*, Anthropic is facing criticism from some core users due to the performance decline of its Claude model. These issues have forced the company to address external speculation about its motives and service capabilities as it reportedly prepares for an IPO. Many developers and heavy users have noted a significant recent drop in the performance of Anthropic's Claude model, manifesting as difficulty following instructions, taking inappropriate shortcuts in complex tasks, and an increase in error rates. Most user complaints are centered around Anthropic's programming tool, Claude Code. The problems appear to be related to recent adjustments Anthropic made to the model's operational settings. The company lowered the default "effort level" to reduce the number of tokens processed per request, aiming to conserve computational resources. Speculation suggests that Anthropic may be facing computational resource constraints following a recent surge in product usage. Anthropic has not provided a detailed response to the complaints. Claude Code's product lead, Boris Cherny, responded to user concerns online, stating that the company adjusted the default "effort level" to "medium" in response to previous feedback about excessive token consumption. However, some users believe the company did not adequately disclose this change.