Apple is trying to lower one of the practical barriers to AI app experiments: usage cost. TechCrunch reports that Apple is waiving cloud API costs for developers with fewer than 2 million first-time App Store downloads, a threshold that would cover many independent teams and smaller software businesses.

The pricing detail matters because Apple is also asking developers to build more directly on Apple Intelligence. Apple’s developer site says apps can now access the on-device foundation model at the core of Apple Intelligence through the Foundation Models framework. The public documentation describes a text-focused on-device large language model for language understanding, structured output, and tool calling.

That makes the program less like a standalone model launch and more like developer infrastructure. Apple is pairing local inference, Swift-native APIs, guided generation, and app-specific tools with a reported cloud-cost break for developers that have not yet reached large App Store scale.

The conservative read is that Apple is still playing to its platform strengths rather than matching the open-ended cloud model strategy of OpenAI, Anthropic, or Google. Developers get a tighter path into system-level AI features, but model availability still depends on Apple Intelligence support and user settings. For smaller app makers, the notable change is that experimenting with Apple’s AI stack may now carry less infrastructure risk while staying close to native iOS, iPadOS, and macOS workflows.