Cursor releases Developer Habits Report on how AI is reshaping software development

Details

• Cursor has published the Cursor Developer Habits Report, summarizing how software development workflows are changing with AI.
• The report is based on what Cursor describes as the most comprehensive dataset on AI coding across major model families.
• Cursor identifies five main themes: developer acceleration, the economics of intelligence, the power user gap, the rise of context, and the shift to automation.
• The research finds developers are increasingly using AI agents to produce larger, "mega" pull requests, signaling deeper automation of coding tasks.
• Cost analysis shows roughly a 7x variation in cost per accepted line of code across different AI model families, highlighting significant efficiency differences.
• Usage distribution is skewed, with power users responsible for a large and growing share of total AI coding activity.
• As AI agents consume more context, input tokens now make up the majority of price-equivalent token spend, shifting optimization focus from outputs to prompts and context windows.
• Cursor provides public access to the full Developer Habits Report via its website for further methodology and data details.

Impact

The Cursor Developer Habits Report gives rare quantitative detail on how AI coding tools are altering engineering economics and workflows. By spotlighting cost-per-accepted-line differences, the growing dominance of power users, and the rising cost of context, it arms teams and vendors with benchmarks that may shape model selection, pricing strategies, and broader platform competition in AI-assisted software development.