Anthropic’s compute deal with xAI has led to an immediate benefit to Claude’s subscribers.
Anthropic has announced that its partnership with SpaceX — the launch and infrastructure arm of Elon Musk’s empire — will give it access to all of the compute capacity at Colossus 1, SpaceX’s flagship AI supercomputer. The result is over 300 megawatts of new capacity, backed by more than 220,000 NVIDIA GPUs, coming online within the month. And unlike many infrastructure deals that promise future benefits, Anthropic has moved immediately to pass the gains to its users.
What’s Changing For Subscribers
Effective today, Anthropic is doubling Claude Code’s five-hour rate limits for Pro, Max, Team, and seat-based Enterprise plans. It is also removing the peak-hours limit reduction on Claude Code for Pro and Max accounts — a restriction that had been forcing users to ration their most intensive sessions to off-peak windows. API rate limits for Claude Opus models are also being raised substantially.
The changes are a direct reversal of a painful squeeze that Anthropic had been putting its users through. Claude’s daily active users on iOS and Android had grown roughly 5x since October 2025, and the infrastructure simply couldn’t keep pace. Anthropic had been forced to introduce weekly rate limits for Pro and Max, tighten session windows during peak hours, and cut third-party tools like OpenClaw off from subscription access entirely. Sam Altman had been publicly gleeful about Anthropic’s resource crunch, with his “ok boomer” quip going viral after Anthropic pulled Claude Code from its Pro plan for a subset of new signups.
Today’s announcement is Anthropic’s answer to all of that.
An Infrastructure Blitz
The SpaceX deal is just one piece of a much larger compute build-out that Anthropic has been assembling at speed. The full picture is striking in its scale:
- A 5 GW agreement with Google and Broadcom for next-generation TPU capacity, coming online from 2027
- A deal with Amazon that includes nearly 1 GW of new capacity by the end of 2026, with additional inference capacity in Asia and Europe
- A strategic partnership with Microsoft and NVIDIA worth $30 billion of Azure capacity
- A $50 billion commitment to American AI infrastructure through Fluidstack
Anthropic already runs Claude across AWS Trainium, Google TPUs, and NVIDIA GPUs — a diversified hardware strategy that gives it both performance flexibility and supply chain resilience. The SpaceX deal adds a fourth dimension: raw scale, fast.
Why This Matters
The timing is not incidental. Anthropic’s ARR has reportedly hit $44 billion, up from $9 billion at the end of 2025, making it arguably the fastest-growing company in technology history. Claude Code alone has crossed $2.5 billion in annualized revenue, with weekly active users doubling since January. At that pace of growth, compute is not just an operational concern — it is the ceiling on revenue. Every user who hits a rate limit is a user who might be considering a switch to Codex or Gemini.
The Colossus deal also carries a longer-range dimension. Anthropic has expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute — effectively moving AI infrastructure beyond terrestrial constraints of land, power, and cooling. It is an ambitious concept, but SpaceX is the only organization with the launch economics to make it a near-term engineering question rather than a research fantasy.
For now, subscribers will be paying attention to something more immediate: a doubled rate limit and no more peak-hour throttling. After months of tightening, that is a meaningful improvement — and a signal that Anthropic’s infrastructure ambitions are finally catching up with its growth.