Cursor Targets $2 Billion Raise at $50 Billion Valuation as Revenue and Margins Inflect

New updates have been reported about Cursor.

AI coding startup Cursor is in talks to secure at least $2 billion in new capital at a pre-money valuation of about $50 billion, according to multiple sources, positioning the four-year-old company as one of the highest-valued private players in developer-focused AI. Existing backers Thrive Capital and Andreessen Horowitz are expected to lead the oversubscribed round, with Battery Ventures and strategic investor Nvidia also weighing participation, though final terms may still shift before closing.

If completed, the deal would almost double Cursor’s prior $29.3 billion post-money valuation from six months ago and effectively price in aggressive growth expectations. Cursor reportedly hit a $2 billion annualized revenue run rate in February and is forecasting more than $6 billion in ARR by the end of 2026, implying at least a tripling of revenue within roughly 10 months despite intensifying competition from rival AI coding products.

A key driver of Cursor’s financial trajectory is its recent shift in unit economics, moving from negative to slightly positive gross margins after it launched its proprietary Composer model in November. By routing workloads through Composer and selectively tapping lower-cost third-party models, including China-based Kimi, Cursor has improved economics on large enterprise contracts, even as it still incurs losses on individual developer accounts.

This pivot toward proprietary technology is also strategic: reducing dependence on external model providers helps Cursor defend against the risk of being displaced by its own suppliers, particularly Anthropic, whose Claude Code product is a primary competitor. The company, founded in 2022 at MIT by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger, declined to comment on the prospective financing, but the size, valuation, and involvement of marquee investors underscore investor conviction that Cursor can become a dominant infrastructure layer in AI-assisted software development.