SpaceX inks compute deal with Reflection AI, an open source AI lab

Overview

SpaceX has finalized a substantial compute agreement with Reflection AI, an open-source AI research lab. Starting July 1, 2026, through 2029, Reflection AI will pay $150 million monthly for immediate access to Nvidia's cutting-edge GB300 AI chips and supporting hardware, all hosted at SpaceX's Colossus 2 data center near Memphis, Tennessee.

Industry Impact

This deal marks a significant shift in AI infrastructure. SpaceX diversifies its offerings, leveraging its extensive infrastructure to become a major AI compute host, challenging traditional cloud providers. For Reflection AI, this secures critical, high-performance Nvidia hardware, promising accelerated open-source AI development. The agreement also solidifies Nvidia's central role in powering frontier AI advancements, emphasizing the growing demand for specialized hardware.

Why It Matters

This partnership vividly demonstrates the escalating "AI arms race" and the immense capital investment required for cutting-edge AI development. Access to scalable, powerful compute is now a decisive strategic advantage. It also highlights how companies with vast physical infrastructure, like SpaceX, are finding new lucrative avenues by meeting the insatiable demand for AI processing power. This deal underscores that the future of AI will be shaped by those who control or can access the most potent compute resources.

Key Points

• Reflection AI to pay $150M/month (totaling $5.4B+) for AI compute from 2026-2029.
• Access to Nvidia's latest GB300 AI chips and supporting hardware.
• Compute resources located at SpaceX's Colossus 2 data center.
• SpaceX strategically enters the high-performance AI compute hosting market.
• Highlights the enormous cost and strategic importance of AI infrastructure for advanced, open-source AI development.