Anthropic's recently launched AI model, Claude Opus 4.7, is currently experiencing significant user backlash. While Anthropic asserted that the new model should feel "more intelligent, agentic, and precise," many users are reporting a different experience.
Critics across social media platforms like X (formerly Twitter) and Reddit claim that Opus 4.7 frequently makes mistakes, exhibits a "combative" demeanor, and consumes an excessive number of tokens, leading to higher operational costs. This has led some to label the model as "dumb" and prohibitively expensive for its perceived output.
Conversely, not all feedback has been negative. Some users argue that the advanced capabilities offered by Opus 4.7 justify its associated costs. Notably, Y Combinator CEO Garry Tan has expressed appreciation for the new model, indicating a mixed reception within the tech community.
This widespread user dissatisfaction, often termed the "Claude-lash," is a relatively uncommon occurrence for a high-profile AI product. It highlights the growing scrutiny on AI model performance, cost-efficiency, and user experience as the technology rapidly evolves.