Last updated: September 24, 2026. OpenAppleModels was built from the ground up on a fundamental commitment: all intelligence remains strictly on the user's device.
OpenAppleModels contains zero tracking code, zero analytics SDKs, and zero telemetry beacons. Neither prompts, tool parameters, world-state data, game dialogue, nor model completions are ever transmitted to SpaceCorps or any third-party cloud service.
The framework interfaces directly with Apple's native on-device FoundationModels system service, running inference on the device. There are no cloud inference endpoints, no remote LLM servers, and no external token meters.
Because computation runs locally on the device hardware, OpenAppleModels requires no API keys, developer tokens, credit cards, or account credentials. You never have to worry about credential leakage or billing surprise.
This static documentation website is hosted on GitHub Pages. It uses no tracking cookies, analytics scripts, or user profiling. Access logs are maintained by GitHub according to GitHub's standard Privacy Policy.