Anthropic has announced a major boost to its Claude Code service, along with a new partnership with SpaceX. The deal involves using SpaceX’s powerful data center in Memphis, Tennessee, to increase compute capacity for Anthropic’s AI models. This move aims to meet growing demand and improve service for its users.

Enhanced Usage Limits and New Capacity

At the recent Code with Claude developer conference, Anthropic revealed that it has doubled the usage limits for its Pro and Max plan subscribers. Specifically, the company increased the five-hour window limits and removed peak-hour restrictions for Claude Code users. Additionally, API limits for its Opus model were raised, providing more flexibility for developers and businesses relying on their AI tools.

The partnership with SpaceX is central to this expansion. Anthropic now has access to over 300 megawatts of new compute capacity, thanks to SpaceX’s Colossus 1 supercomputer. This system features more than 220,000 NVIDIA GPUs, including cutting-edge H100, H200, and upcoming GB200 accelerators. These are among the most powerful AI training tools available today.

SpaceX’s Supercomputer and Future Plans

SpaceX highlighted Colossus 1’s capabilities, emphasizing its dense GPU deployment. The company also expressed interest in building orbital data centers, which could provide even more compute power in space. The idea is to develop “multiple gigawatts” of orbital compute capacity, addressing the increasing need for faster, more scalable AI training.

This partnership marks a notable shift, especially considering Elon Musk’s previous criticisms of Anthropic. In early 2024, Musk publicly questioned the company’s practices and ethics. However, after recent discussions, Musk shared that he was impressed by Anthropic’s efforts to ensure Claude’s safety and usefulness for humanity. This change signals a new chapter for the two companies.

Growing Demand and Industry Impact

Demand for Claude Code and similar AI products has surged recently. Factors include users moving away from OpenAI due to controversies related to military contracts, and a shift towards more complex multi-agent workflows in software development. As a result, Anthropic has faced challenges in keeping up with this demand, leading to temporary usage limits and outages.

To address this, the company has secured deals with major players like Microsoft, Google, Amazon, and Nvidia. These partnerships are expected to enhance Anthropic’s compute infrastructure over time, though some benefits will take months to fully materialize. The SpaceX deal is a key part of this broader effort to scale up capacity quickly.

Many developers have voiced frustration over usage restrictions, especially on platforms like Hacker News and Reddit. They want more access to powerful AI tools for their projects. Anthropic’s recent moves aim to balance demand with available resources, ensuring that users can leverage Claude Code more effectively in the future.

This strategic expansion shows how AI companies are racing to meet the rising appetite for advanced models. By teaming up with SpaceX and securing large compute deals, Anthropic is positioning itself to stay competitive in the rapidly evolving AI landscape. The next few months will reveal how well these efforts translate into better service and broader adoption.

Inspired by

• https://arstechnica.com/ai/2026/05/anthropic-raises-claude-code-usage-limits-credits-new-deal-with-spacex/

Sources

• anthropic.com
• x.ai