Cursor Hits $2B Revenue Run Rate
AI coding assistant Cursor reportedly surpassed $2 billion in annualized revenue, signaling explosive growth in AI developer tools. Stripe also released a preview feature designed to help businesses turn their AI costs into profit centers. OpenAI's latest Codex release includes features for branching work into sub-agents and improved real-time voice sessions.
Cursor's rapid ascent is underscored by its revenue doubling in just the last three months, a significant acceleration for the four-year-old startup. This growth counters recent social media discussions questioning the company's momentum as some individual developers switched to competing tools like Anthropic's Claude Code. The company's financial strength is largely driven by a strategic pivot towards larger companies. Enterprise customers now represent approximately 60% of Cursor's revenue, a shift from its initial focus on individual developers. This move provides more stable, larger contracts to offset any churn from smaller users. This revenue milestone follows a massive $2.3 billion funding round in November 2025, which valued Cursor at $29.3 billion. The investment was co-led by Accel and Coatue, with participation from major tech players including Google and Nvidia. The market for AI developer tools is expanding rapidly, with a projected compound annual growth rate (CAGR) of over 26% from 2025 to 2030. This growth reflects a broader industry trend where AI coding assistants are becoming essential infrastructure for software development teams. Stripe's new feature directly addresses a major pain point for businesses building on AI: the variable and often high cost of using models from providers like OpenAI and Google. The tool allows companies to automatically add a profit margin on top of the raw token costs they incur, turning an operational expense into a direct revenue stream. OpenAI's update to Codex introduces more sophisticated workflow management. The ability to "fork a thread into sub-agents" allows a single AI assistant to delegate and orchestrate multiple specialized agents in parallel. This is designed for complex tasks like comprehensive code reviews or implementing multi-step features. This multi-agent capability is a core focus for OpenAI, positioning Codex as an orchestration layer for AI-driven development rather than just a code completion tool. The approach enables developers to manage long-running, parallel tasks from a central command center, with agents working in the background across an entire codebase.
