Anthropic’s Fearless Expansion into the Future of AI
Anthropic announced a series of infrastructure deals and product changes, headlined by a partnership with SpaceX that gives the company access to one of the world's largest AI data centers. The moves significantly expand Anthropic's compute capacity at a time when demand for Claude, particularly the Claude Code coding tool, has been straining the company's resources.
The SpaceX deal: Colossus 1
The centerpiece is a deal with SpaceX (which merged with xAI earlier this year) to use the full compute capacity of the Colossus 1 data center in Memphis, Tennessee. That means access to over 220,000 NVIDIA GPUs, including H100, H200, and GB200 accelerators, adding more than 300 megawatts of new capacity within the month.
The partnership is notable given the history between the two sides. Elon Musk had previously called Anthropic "misanthropic and evil," but said in a post on X that he spent time with senior Anthropic staff last week and came away "impressed." He added that SpaceXAI had already moved its own training workloads to Colossus 2, freeing up the original facility. Musk also noted he reserves the right to reclaim compute if Anthropic's AI "engages in actions that harm humanity."
Anthropic has also expressed interest in partnering with SpaceX to develop orbital AI compute capacity. This remains an early-stage idea, not a concrete commitment.
What changes for users
The new capacity translates into three immediate product changes:
Claude Code's five-hour rate limits are doubled for Pro, Max, Team, and seat-based Enterprise plans. Peak-hour usage restrictions on Claude Code are removed for Pro and Max accounts. API rate limits for Claude Opus models are raised substantially. Tier 1 API limits saw a 1,500% increase in maximum input tokens per minute and a 900% increase in output tokens per minute, enabling businesses to integrate the model at far greater scale.
A broader compute strategy
The SpaceX deal is one piece of a larger infrastructure push. Anthropic's other recent agreements include an up to 5 GW deal with Amazon, with nearly 1 GW of new capacity expected by the end of 2026; a 5 GW agreement with Google and Broadcom, set to begin coming online in 2027; a strategic partnership with Microsoft and NVIDIA that includes $30 billion of Azure capacity; and a $50 billion investment in American AI infrastructure with FluidStack.
Most of this capacity is still months or years away from being operational. The combined commitments total over 10 GW, but only the SpaceX and Amazon allocations are near-term.
Anthropic also announced plans to expand into Europe and Asia through its Amazon collaboration, targeting enterprise clients in regulated sectors like finance, healthcare, and government who need to store and process data locally.
A deal forged under pressure
The deal arrives at a complicated moment for Anthropic. In March, the Pentagon designated the company a supply chain risk, an unprecedented action against a U.S. company, after Anthropic refused to grant the Department of Defense unrestricted access to Claude for all purposes, insisting its models should not be used for fully autonomous weapons or domestic mass surveillance. A federal judge in San Francisco subsequently blocked the Pentagon's actions, calling them a violation of Anthropic's constitutional rights. The litigation is ongoing.
Anthropic is no longer playing the role of a technology leader without scale. The SpaceX deal, combined with over 10 GW of future capacity from Amazon, Google, and Microsoft, puts the company in a position to compete directly with OpenAI. The question is no longer whether Claude is a good enough model, but whether Anthropic can build out its infrastructure fast enough to turn its technological edge into a market one.
Andrey Hristov