Musk sells Anthropic AI compute: $1.25 billion a month
Anthropic is paying xAI, Elon Musk's company, $1.25 billion a month for computing power. The unexpected deal exposes how fiercely AI companies are competing for GPU resources. Compute has become the scarcest asset for building competitive next-generation models.
AI-processed from TechCrunch; edited by Hamidun News
Anthropic has agreed to pay Elon Musk's company $1.25 billion a month for compute to run neural networks. The deal, which few in the industry saw coming, shows how fiercely AI companies are competing for compute, the scarcest resource in today's artificial intelligence market.
Why is Anthropic paying xAI this much?
Compute is the scarcest resource in the AI market. Anthropic is competing for access to the powerful computing clusters needed to train large language models. xAI owns Colossus, one of the most powerful clusters in the world, with 100,000 H100 GPUs.
What is Colossus and why does it matter?
Colossus is xAI's computing cluster of 100,000 H100 GPUs. These graphics processors are the foundation for training large language models. A single H100 costs roughly $40,000, which gives a sense of the infrastructure's scale.
- Nov 2, 2026LLM Agent Built a Gearbox in KOMPAS-3D, but Scored Only 50 out of 100 on the Specification
- Nov 2, 2026Agents ignored the prompt to read memory in 7 of 7 sessions: an analysis of 13 thousand logs
- Nov 2, 2026Karpathy Built HackerTube, a Catalog of 6,454 Videos, in Two Hours with Claude Code
- Nov 2, 2026We Audited 929 Automations at Alpina: What Corporate AI Training Should Look Like in 2026
Key Questions Answered
Why is Anthropic paying xAI this much?
Compute is the scarcest resource in the AI market. Anthropic is competing for access to the powerful computing clusters needed to train large language models. xAI owns Colossus, one of the most powerful clusters in the world, with 100,000 H100 GPUs.
What is Colossus and why does it matter?
Colossus is xAI's computing cluster of 100,000 H100 GPUs. These graphics processors are the foundation for training large language models. A single H100 costs roughly $40,000, which gives a sense of the infrastructure's scale.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.