Anthropic has formed a partnership with SpaceX, enabling the AI firm to access xAI's Colossus 1 facility. This agreement prompts Anthropic to increase usage thresholds for subscribers to Claude Code and relax additional constraints, according to an official announcement.
Under the terms, Anthropic gains full access to the computing resources at the Colossus 1 site operated by SpaceX, which plans to expand by over 300 megawatts of power in the coming weeks. The company has also secured similar arrangements with Amazon and Google in recent times. Furthermore, the partners have voiced intentions to collaborate on building several gigawatts of AI processing power in space.
SpaceX has submitted documents to the FCC outlining plans to deploy one million satellites aimed at establishing a data center in orbit.
Subscribers on paid tiers, including Pro, Max, Team, and Enterprise, will experience prompt improvements from this development. Anthropic is raising the five-hour usage caps for Claude Code by 100 percent. The firm is also eliminating peak-time limitations for Pro and Max accounts while significantly elevating API thresholds for the Claude Opus variants.
For context, recent interactions with Anthropic's leadership revealed a strong focus on aligning Claude's development with positive human impact, leaving a positive impression of their expertise and commitment.
The alliance between Anthropic and Elon Musk's SpaceX, which has acquired xAI, might appear unexpected. Musk had previously labeled Anthropic as 'misanthropic and evil' and voiced ongoing concerns about its leader, Dario Amodei. However, in a recent X post, he expressed admiration for the Anthropic staff and suggested that Claude is likely to prove beneficial. In a follow-up message, he noted that they could retrieve the computing resources if the AI system causes harm to people.