Cursor

SpaceX, Elon Musk’s aerospace company, Indian on Wednesday that he had reached an agreement with AI startup Cursor. According to this agreement, this company could be acquired for 60,000 million dollars. If everything is confirmed, xAI will finally have a programming AI agent with which to compete. Claude Code, Codex or Gemini AI. Restructuring. Elon Musk posted a message on X in March in which claimed that “xAI was not created the right way initially, so it’s rebuilding from the ground up.” The company’s trajectory has been erratic and most of its original founders ended up leaving the company in recent months, but for that restructuring Musk hired Andrew Milich and Jason Ginsberg, two of the co-founders of Cursor. And that has been the trigger for this agreement. Cursor’s rating skyrocketed. By November 2025, Cursor was the undisputed leader in the programming AI agent segment, with $3.4 billion in funding and a fantastic reputation among developers. It had reached $100 million in annual recurring revenue in less than two years, a figure that few startups of its generation can aspire to. The Claude Code earthquake. The arrival of Claude Code changed everything thanks to his extraordinary behavior and its integration with Anthropic models, and while OpenAI also promoted Codex and the era of vibe-coding made all these tools gain more strength than ever. However, for Cursor these launches were problematic because its competitors could work directly with companies because they had something that Cursor did not: computing capacity. Cursor needed an ally. In the statement of the agreement, those responsible for Cursor they explain that the lack of access to computing power to train their own AI models had been a major bottleneck for its growth. Both Anthropic and OpenAI have access to several present and future GW of compute thanks to their agreements with hyperscalers (Amazon, Google, Microsoft). But no matter how good Cursor was, it was competing with giants with many more resources. The agreement with SpaceX gives it access to xAI’s AI supercomputer, which is precisely perfect for training LLMs and which according to its managers allows them to “drastically scale the intelligence of our models”: But xAI too. On the other side of the agreement there is also another winner. xAI have their Colossus supercomputing cluster, access to SpaceX resources (of which it is part) and a significant base of users (and their data) thanks to Grok. What it does not have is a product that competes with Claude Code or Codex, and attempts to develop it have been unsuccessful. Buying Cursor solves that problem at once: instead of working on building a product for years while its competitors continue to advance, xAI directly integrates the team that already had that product and also adds the users who were already using Cursor. Will it be enough? The question this agreement must answer is whether it is enough to put xAI on the real map of artificial intelligence. The company has a minor presence in this market despite the efforts of Elon Musk, and although it will now have a product respected and valued by users, it will be interesting if that is enough to compete with its rivals in this area. The IPO as a contextual framework. SpaceX has been preparing to go public for months, and it is expected that this will be one of the largest Public Sale Offers (IPO) of history. Acquiring Cursor before or after that deal has clear financial implications because SpaceX has two options. You can buy Cursor for $60 billion, or simply pay $10 billion for a close collaboration agreement that does not include the acquisition. SpaceX will make that decision before the end of the year, but this agreement seems to suit both it and Cursor very well. Image | Gage Skidmore In Xataka | Elon Musk knows that TSMC is overwhelmed: Terafab is his idea to completely change the global chip industry