Apple ships new Mac mini and Mac Studio as local-AI alternative to per-token cloud costs

Apple's pitch reframes AI compute as a capex-versus-opex decision, targeting businesses that balk at metered token billing from cloud model providers. It matters less as a hardware story than as a signal that inference economics are shifting toward on-premise deployment, which could pressure per-token pricing models if buyers treat local hardware as a substitute. The open question is whether businesses accept owning and maintaining that capacity over the flexibility of renting it.