Cursor, the AI-powered coding assistant platform, is facing intense backlash after it quietly restructured its Pro subscription plan, replacing the previous 500-request-per-month model with a token-based system. The change, which was rolled out with limited communication, drastically reduced usage quotas and left many users shocked by surprise overage charges.

Many developers reported quickly burning through their newly issued token quotas, with one team allegedly exhausting a $7,000 annual plan in just a single day. Social media platforms were soon flooded with posts and threads criticizing the lack of notice and transparency. Frustrated users began exploring alternatives such as Claude Code and other emerging AI coding tools.

In response to the uproar, Cursor published an official blog admitting they had "missed the mark" on communication and promised to issue refunds for unexpected charges. They also acknowledged the strain caused by shifting economics as more capable, compute-intensive AI models drive up costs.

According to Cursor, the new pricing model was a necessary business move to ensure long-term sustainability. However, the episode has become a cautionary tale in how not to roll out major product changes. While the company may recover from the PR damage, the incident underscores how pricing changes in AI SaaS are now a minefield — any misstep can push users toward rival platforms.

This situation illustrates the larger tension in the AI tooling space: balancing innovation and cost transparency. As Cursor recalibrates its policies, its user base remains wary, and the developer community will be watching closely for its next move.