Anthropic's partnership with SpaceX, announced on 7 May, represents one of the more unexpected infrastructure deals in the AI industry. By accessing xAI's Colossus 1 supercomputer — originally built by Elon Musk's AI venture — Anthropic adds over 300 megawatts of computing capacity, equivalent to more than 220,000 Nvidia H100 GPUs, to its existing infrastructure.
The immediate impact for developers was tangible: Claude Code rate limits for Pro, Max, and Team subscribers doubled overnight, and peak-hour usage restrictions were eliminated entirely. API rate limits for Opus-tier models also increased. For context engineers who rely on Claude Code for production development workflows, this removes one of the most common friction points — hitting rate limits during intensive coding sessions.
The deal also includes what Anthropic described as an 'orbital compute partnership in development', suggesting future satellite-based inference or training infrastructure, though no details were provided.
This partnership sits alongside Anthropic's broader compute expansion strategy. The company also announced a $200 billion Google Cloud contract, a Broadcom partnership for next-generation TPU capacity, and an $1.8 billion Akamai computing deal — all within the same week. Anthropic's run-rate revenue now exceeds $30 billion, up from approximately $9 billion at end of 2025, with over 1,000 business customers each spending more than $1 million annually. CEO Dario Amodei described Q1 2026 as delivering '80x year-over-year growth in revenue and usage'. Whether the Colossus partnership signals a broader thaw between the Anthropic and Musk camps, or is purely a capacity-driven commercial arrangement, remains to be seen.
