Apple waives cloud API fees for developers under 2M downloads using Private Cloud Compute

Apple will eliminate cloud API costs for developers with fewer than 2 million first-time App Store downloads who use its Foundation Models running in Private Cloud Compute, the company announced at its Worldwide Developers Conference on Monday.

The move targets indie developers similarly to Apple's Small Business Program, which offers reduced commission rates to smaller app creators. "It's access to frontier-tier level intelligence with unparalleled privacy protections, because getting started exploring ideas shouldn't be held back by infrastructure costs," according to Apple's presentation.

Framework expansion

Apple's Foundation Models framework now includes image input support and integration with third-party cloud model providers. The API can connect to developers' chosen cloud model provider for more complex tasks that require larger models beyond Apple's own infrastructure.

The company positioned this as making large cloud models "as accessible as possible" for developers who need additional capabilities.

Industry context

The announcement reflects mounting pressure on AI infrastructure costs across the industry. Meta and Amazon have both discontinued internal AI token usage leaderboards where developers previously competed on AI tool experimentation. Uber recently disclosed it consumed its entire 2026 AI budget within four months.

Pricing details for developers exceeding the 2 million download threshold were not disclosed. Apple did not specify which Foundation Models are included in the program or provide benchmark scores for its models.

What this means

Apple is using its ecosystem leverage to subsidize AI experimentation for smaller developers, betting that reducing friction will drive adoption of its AI infrastructure. The 2 million download threshold creates a clear on-ramp for indie developers while maintaining revenue from larger studios. By opening the framework to third-party cloud providers, Apple acknowledges its models alone won't meet all developer needs—a pragmatic stance that contrasts with more closed AI platforms. The timing aligns with broader industry cost-cutting around AI experimentation, suggesting even large players are reassessing unsustainable burn rates.