SpaceX Grants Anthropic Access to Colossus 1 Supercomputer for AI Model Training

SpaceX has signed an agreement with Anthropic, allowing the AI company to use its Colossus 1 supercomputer for running AI models. This partnership comes as Elon Musk, SpaceX CEO, is involved in a legal dispute with OpenAI, Anthropic’s main competitor. Anthropic is also a direct rival to Musk’s own AI startup, xAI.

Key Highlights

• SpaceX allows Anthropic to use Colossus 1 supercomputer for AI model training.
• Colossus 1 features over 220,000 NVIDIA GPUs including H100, H200, and GB200 accelerators.
• Anthropic doubles Claude Code limits and removes peak hour restrictions after gaining more compute power.
• Elon Musk says Anthropic's team convinced him of their ethical approach to AI development.

Details of the SpaceX-Anthropic Agreement

Anthropic will use Colossus 1, which SpaceX describes as one of the world’s largest and fastest-deployed AI supercomputers. The system is designed for AI training, fine-tuning, inference, and other high-performance computing tasks. Colossus 1 contains over 220,000 NVIDIA GPUs, including H100, H200, and GB200 accelerators.

Elon Musk commented on the agreement through a post on X. He said he met with Anthropic’s senior officials to understand their approach to AI safety. Musk stated that he was impressed by their commitment to ethical practices. He added that none of the Anthropic team set off his “evil detector” and expressed confidence that their Claude AI model would be safe for humanity.

Musk explained that SpaceXAI had already moved its own training to Colossus 2, making Colossus 1 available for Anthropic’s use. This move comes as Anthropic faces challenges with compute resources, which had previously led to reduced usage limits for its users.

Industry and Public Reactions

The agreement has generated significant discussion online. Many users on X speculated that Musk’s decision may be aimed at helping Anthropic compete with OpenAI. Memes and comments suggested that Musk’s main goal could be to see OpenAI fall behind in the AI race.

Anthropic is seen as having an advantage over OpenAI in the enterprise sector. However, limited compute resources had forced Anthropic to restrict usage for its Claude Code product. OpenAI has tried to capitalize on this by promoting its own coding agent, Codex.

Impact on Anthropic’s Services

With access to Colossus 1, Anthropic quickly updated its user limits. The company doubled Claude Code’s five-hour rate limits for Pro, Max, Team, and seat-based Enterprise plans. It also removed peak-hour usage reductions for Pro and Max accounts. Additionally, Anthropic announced substantial increases in API rate limits for its Opus models.

Anthropic has also expressed interest in working with SpaceX to develop multiple gigawatts of orbital AI compute capacity. Elon Musk has promoted the idea of space-based data centers, which could provide sustainable power and reduce environmental impact on Earth.