SpaceX Secures $60B Option or $10B Deal with Cursor for AI Coding Push

SpaceX revealed a partnership with Cursor on April 22, 2026, enabling joint development of advanced AI for coding and knowledge work. The agreement includes SpaceX's right to acquire Cursor for $60 billion later in 2026 or compensate $10 billion if no acquisition occurs. Cursor gains access to xAI's Colossus supercomputer, equivalent to one million Nvidia H100 GPUs, to train its Composer models, including the upcoming Composer 2.5. Prior collaborations involved Cursor training on xAI infrastructure using tens of thousands of GPUs, building on hires of former Cursor engineers by xAI.

Cursor, developed by Anysphere and founded by MIT graduates Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger, offers an AI-powered code editor that integrates intelligence directly into development workflows. The platform supports agentic coding agents via Cursor 3 and automates tasks like code rewriting. This deal addresses Cursor's compute constraints, allowing independence from rivals like OpenAI and Anthropic, whose models it previously relied on. SpaceX aims to combine Cursor's engineer distribution with its compute power to challenge leaders in AI coding services.

The partnership holds strategic importance amid SpaceX's AI expansion, following its acquisition of xAI and IPO filing. It positions SpaceX to compete with Anthropic's Claude Code and OpenAI's Codex by creating superior models for software engineering. Industry observers note it fills gaps for both: Cursor secures massive compute, while SpaceX gains coding expertise and distribution. Discussions with Mistral for a three-way tie-up were also explored, enhancing competitive edge in the growing vibe coding market.

For Cursor, valued recently above $50 billion in scrapped funding talks, the arrangement accelerates model scaling after rapid progress from Composer to Composer 2. SpaceX's involvement could integrate Cursor's tech into Grok products, boosting revenue via APIs and supporting its public offering ambitions. This move underscores compute as a key bottleneck in AI development, pressuring competitors' margins.