OnceUponaBookshelf
What does "on-device AI" mean, and why does it matter?
Published 23 August 2026 · Updated 12 October 2026 · Written and fact-checked by the OnceUponaBookshelf team
On-device AI means the model that writes a response runs directly on your phone's own chip, instead of your words being sent over the internet to a company's server and a reply sent back. On recent iPhones, Apple provides this through Apple Intelligence and its Foundation Models framework, which apps can use to generate text without anything leaving the device. For an app that handles information about your child — a first name, an age, their interests, a story idea — that difference matters: on-device processing means the information can stay on the phone rather than being transmitted, logged or stored by a third party. It also isn't universal. It needs a recent iPhone with Apple Intelligence turned on, so a well-designed app should tell you when it's working on-device and when it isn't, and ask before it falls back to the cloud.
What happens when you tap "generate"?
Most AI apps work the same way: you type something, the app sends it over the internet to a company's servers, a large model running on that company's hardware generates a response, and the result is sent back to your phone. That's how ChatGPT, most chatbots, and most AI writing tools work — the intelligence lives in the cloud, and your device is mostly a window onto it.
On-device AI works differently. Recent iPhones can run a language model directly on the phone's own chip, so your prompt never has to leave the device — the thing generating the response is sitting right there in your pocket. Apple makes this available to app developers through its Foundation Models framework, the same on-device model that powers Apple Intelligence.
Why does it matter more for children's apps?
For a lot of AI use cases, it genuinely doesn't matter much whether the processing happens on your device or in the cloud. It matters a great deal more for apps that handle information about a child, because:
- Nothing to intercept. Data that never leaves the device can't be caught in transit, logged by a server, or included in a data breach at a company you've never heard of.
- Nothing to retain by accident. Cloud services vary in how long they keep what you send them. On-device processing sidesteps the question entirely.
- Less to consent to. When nothing is sent, there's no third party to vet — which lines up with the "collect only what you need" principle in the UK ICO's Children's code.
What are the honest limits?
On-device AI isn't free of trade-offs. It needs a fairly recent, fairly capable phone — older devices simply can't run the models involved. And on-device models are smaller and somewhat less capable than the largest cloud models, which is part of why some apps (including ours) fall back to a cloud model on devices that can't run the on-device version, rather than just not working at all.
The honest way to think about it: on-device AI is a meaningfully stronger privacy default where it's available, and a well-designed app should be upfront about when it's using it and when it isn't — not just claim "on-device" as a marketing line while quietly sending everything to a server anyway.
Which iPhones actually support on-device AI?
Last reviewed 12 October 2026, against Apple's Apple Intelligence requirements.
Apple Intelligence runs on these iPhones:
- iPhone 15 Pro and iPhone 15 Pro Max (the standard iPhone 15 and 15 Plus don't qualify)
- iPhone 16 models, including iPhone 16e
- iPhone Air
- iPhone 17 models and later
Apps that write text with Apple's on-device model also need iOS 26 or later, with Apple Intelligence turned on in Settings → Apple Intelligence & Siri, the iPhone and Siri set to the same supported language (English, including UK English, is supported), and enough free space for the model — Apple says up to 8 GB, or up to 14 GB on iPhone 17 Pro, iPhone 17 Pro Max and iPhone Air. Apple adds new iPhones to the list each year, so that Settings screen is the quickest way to check a specific phone.
This is why OnceUponaBookshelf doesn't require on-device AI to work at all. The app runs on any iPhone with iOS 18 or later; on a phone that doesn't meet the bar above, it asks your permission and writes that story in the cloud instead of refusing to write one.
How does OnceUponaBookshelf use it?
OnceUponaBookshelf tries to write every story on-device first, using Apple's on-device model running locally on your iPhone — in that case, your child's name, age, interests and story idea never leave the device. On a phone that can't run the on-device model, the app asks your permission and writes that one story with Cloudflare Workers AI instead, and nothing is kept once it's delivered back to your device. Narration is separate: if you add it, the finished story text is sent to ElevenLabs to make the audio. See exactly how in how it works, or the FAQ.
Sources
- Apple Support — How to get Apple Intelligence
- Apple Developer — Foundation Models framework
- Cloudflare — Your data and Workers AI
Related reading
See on-device AI in action
Story writing is free for everyone. Remove Ads if you'd rather skip the short ad.