Text to speech that works offline
Most read-aloud apps are a thin client for a cloud service: your text goes to a server, audio comes back. That’s why they stop on a plane, stutter in the subway and ask you to sign in. Aloudy puts the voice itself on your iPhone.
How it works
Aloudy’s voices are neural text-to-speech models, the same kind of AI that powers cloud voices, compressed to run on the iPhone’s own chip. The first time you pick a voice, the app downloads its model once (the multilingual voice is about 140 MB). From then on, every sentence is generated on your device, a moment before you hear it.
Text recognition for scanned pages uses the iPhone’s on-device engine too. Your library, reading positions and settings live in the app’s storage on your device.
What works offline
- Reading aloud with any downloaded voice, at any speed
- Every document already in your library: PDF, EPUB, DOCX, TXT, Markdown, HTML, RTF
- Scanning a printed page and listening to it
- Articles and classic books you’ve already added
- Background playback, lock screen and Control Center
What needs a connection
- Downloading a voice for the first time
- Opening a web page or a shared link
- Downloading a book from the classics catalog
Tip: before a flight, download the voice you like and add what you want to read. The rest of the trip is offline.
Why on-device is also private
If the voice runs in the cloud, your text has to travel there. With Aloudy it doesn’t: there is no account, no server that receives your documents and no ads. Contracts, medical letters, drafts and study notes are read on your phone and stay on your phone.
On-device vs cloud text to speech
Questions about offline reading
Is there a text to speech app for iPhone that works without internet?
Yes. Aloudy generates speech on the iPhone itself. After a one-time voice download it reads PDFs, e-books, documents and scanned pages with no internet connection.
Do offline voices sound worse?
Aloudy’s voices are neural models, not the old robotic system voices. They are the same kind of AI voices cloud apps use, running locally.
How much storage do the voices take?
The multilingual voice that reads 31 languages is about 140 MB. English and Chinese voices are downloaded separately when you choose them.
Can I listen with the screen locked?
Yes. Aloudy keeps reading in the background, with controls on the lock screen and in Control Center, and carries on after calls and when you switch headphones.