SpaceX Powers Anthropic with 300 MW of AI Compute as Musk Battles OpenAI

TL;DR: SpaceX will feed Anthropic’s Claude models with 300 MW of compute from its Memphis Colossus 1 facility, a move Elon Musk touts as “good for humanity” while he sues OpenAI.

Context Elon Musk’s lawsuit against OpenAI has sharpened the spotlight on AI competition. In the same week, SpaceX announced a data‑centre agreement with Anthropic, the creator of the Claude chatbot series. The deal arrives as Anthropic rolls out new capacity limits for its paid users.

Key Facts SpaceX’s Colossus 1 centre in Memphis houses more than 220,000 Nvidia processors. Within a month, the facility will deliver 300 megawatts of power to Anthropic—enough electricity for over 300,000 homes. Musk said the meeting with Anthropic’s leadership triggered no “evil detector,” calling their work “good for humanity.”

Anthropic responded by doubling rate limits for its Claude Code service, removing peak‑hour caps for Pro and Max plans, and expanding request volume for Claude Opus models. The company also unveiled a “dreaming” feature that lets AI agents review prior sessions and adjust to user preferences.

What It Means The infusion of SpaceX compute capacity positions Anthropic to scale Claude Pro and Claude Max offerings, potentially narrowing the gap with rivals that rely on larger cloud providers. Musk’s public endorsement may signal a strategic alignment against OpenAI, whose leadership he is currently litigating.

Looking ahead, Anthropic plans to explore gigawatt‑scale orbital data centres with SpaceX—a venture that could redefine AI infrastructure if technical and financial hurdles are cleared. Watch for how this partnership influences market share and regulatory scrutiny in the AI arms race.