Apple Inc. (AAPL  ) is pitching its newest Macs as a way for businesses to run powerful AI models without paying per-token cloud charges, taking aim at the economics behind the AI infrastructure boom. "There's no cost per token. You're just using the machine again and again," Apple hardware chief Johny Srouji told Reuters.

Apple demonstrated four Mac Studios running a trillion-parameter AI model together, with Reuters reporting the cluster could operate from a single wall outlet.

Apple Pushes AI Onto the Mac

In Apple's framing, businesses can pay upfront for the hardware and run AI models locally instead of continually paying cloud providers or AI companies for usage.

That hardware cost is substantial. The M5 Ultra Mac Studio starts at $5,499, while a version with 256GB of memory and 16TB of storage costs $18,299. Apple's 512GB configuration, which can handle even larger models, arrives in late October.

Apple raised Mac prices in June citing surging memory and storage costs.

Jensen Huang Says 'Compute Is Revenue'

Nvidia Corp. (NVDA  ) CEO Jensen Huang has described the AI economy differently.

"Its tokens are productive and profitable. Now, compute is revenue," Huang said in Nvidia's latest earnings release.

Nvidia's Data Center revenue reached $89 billion last quarter, up 117% from a year earlier, as cloud providers and AI companies continued spending heavily on computing infrastructure.

Apple's pitch suggests some inference workloads could shift away from rented data-center capacity and onto local machines.

Nvidia is also pursuing the same shift. Its DGX Spark and RTX Spark products are designed to run increasingly capable AI models locally, making desktop AI another battleground between the companies.

Apple Faces an Enterprise Climb

Prediction-market traders still expect Nvidia to stay on top. Polymarket gives Nvidia a 78% chance of ending 2026 as the world's largest company, compared with 15% for Apple, on about $7.5 million in trading volume.

Apple held just 4.6% of the enterprise PC market compared with 91.3% for Windows, according to IDC data cited by Reuters.

Apple's own most advanced cloud model, meanwhile, reportedly runs on Nvidia GPUs through Google Cloud.