Free iPhone text to speech, and where free stops

The App Store says “Free” at the top of almost every text-to-speech app, and it is telling the truth about exactly one thing: the download. What happens after you paste a chapter into it is decided somewhere else, by a meter the listing mentions in the fourteenth bullet point or not at all.

There are five of these meters in circulation. Knowing which one an app uses tells you more about whether you will actually use it than any amount of voice-quality marketing.

Every one of those numbers is discoverable from the App Store listing in half a minute, and the place to look is not the description.

• Scroll to the Information section at the bottom of the listing and look for In-App Purchases. If the row exists, tap it: Apple prints every product and its price, and that list is the real price sheet. It is also more reliable than the description — Local Narrator’s description says a single $4.99 unlock while its purchase list shows $9.99.
• If there is no In-App Purchases row at all, there is nothing inside the app to buy. On 14 September 2026 that was true of Sandbook, Play Text and the developer-facing Vocalize AI among the free apps, and of KO Reader among the paid ones.
• Check the size. A reader that generates speech on your phone has to carry a model: the ones that do are 250 MB to 650 MB downloads. A 15 MB reader is either using Apple’s built-in voices, fetching a model later, or sending your text to a server.
• Read the free tier’s unit. Characters, hours, voices, books, or access. The unit tells you what will run out and when.

Free. No in-app purchase, no subscription, no character or minute cap. All 41 Kokoro voices are available from first launch — none of them held back as the good ones — and there is no account, so nothing to sign up for and nothing to cancel. That is possible because the whole model runs on your phone: there is no server-side bill for anyone to meter.

The price is paid in features instead, and they are not small ones. Real-time reading pauses when the screen locks, because the model holds real GPU memory and iOS will not let that continue once the app leaves the foreground; only a chapter prepared ahead of time plays locked. No text recognition for image-only PDF files. No DOCX, MOBI or AZW3. No iCloud sync between your iPhone and your iPad. Voices in six languages, not dozens. Audio export is M4A from the Reader tab only, so a whole book cannot be turned into an audio file.

If any one of those is the thing you were going to use it for, pay for one of the apps above; they are good, and they are better at exactly those jobs. The free text-to-speech page lays out the same trade from the product side, and the alternatives page puts Sandbook next to the paid readers row by row.

Before installing anything: iOS has Speak Screen built into Accessibility, it costs nothing, it uses Apple’s own voices and it is one swipe away. For hearing an email or a short web page it is genuinely enough, and no app in this article replaces it. Turning it on takes a minute; the steps are in our guide to text to speech on iPhone.

What it does not have is a library. There is no import of a 200-page PDF as one listening item, no per-document position to come back to, no read-along highlighting that follows a book, and no way to keep a shelf of things you are part-way through. Speak Screen compared with Sandbook goes through that difference properly.