OpenAI and rival AI labs are buying tens of thousands of Mac minis to train computer-use agents

The business move

OpenAI has bought tens of thousands of Mac minis and Mac Studios to build and train AI agents designed for computer use. Rival AI company Anthropic is also relying heavily on Apple hardware. The demand for these machines is so strong that the highest-end Mac models have been consistently sold out for months. This surge in enterprise purchasing coincides with Apple’s reported 29 percent jump in Mac revenue, reaching $10.4 billion in the June quarter.

Why it matters

Mac minis and Mac Studios offer a unique hardware combination that appeals to AI labs focused on agent training and software development. Unlike standard cloud GPUs or custom AI chips, Apple machines provide a powerful, energy-efficient, and scalable platform optimized for macOS-based workflows. This hardware preference signals a shift in how certain AI builders are structuring their compute infrastructure, blending desktop-style power with specialized agent software demands.

The high demand and persistent supply constraints mean AI labs must plan hardware procurement far in advance. For Apple, this translates directly into stronger Mac sales driven by a niche yet rapidly growing customer segment beyond traditional creative professionals. Companies provisioning these Macs face higher costs and potential delays but get optimized performance for the kinds of user-interactive AI agents being developed.

Who gains and who gets squeezed

Apple solidifies a new foothold in the AI hardware market. Its rising Mac sales clearly reflect growing adoption beyond usual creative and software development customers. AI startups and labs that secure these devices can run their agent training and inference closer to the end user environment, potentially improving speed and workflow integration.

Cloud GPU providers and traditional AI chip vendors encounter added competition, especially for workloads that are not purely raw compute but require tight integration with software agents controlling desktop applications. Meanwhile, organizations waiting for Macs face longer lead times, raising operational risks and capital outlay.

What to watch next

Track supply availability and price movements for Mac minis and Mac Studios as AI demand grows. Watch if other AI firms shift to or away from Apple hardware for agent training, which could indicate changes in compute infrastructure strategies. Apple’s ability to expand its AI-related hardware sales could pressure cloud and chip incumbents and reshape AI workstation dynamics. Lastly, follow how this hardware choice impacts development velocity and agent performance in real-world deployments.

AI Quick Briefs Editorial Desk