How to listen to research papers on iPhone

Most text to speech apps read a research paper the way the PDF stores it, not the way you read it. Two columns get mixed line by line, every "[12]" and "(Smith et al., 2020)" is read out, and the last ten minutes are the reference list. Narratorr reads papers in reading order and leaves the citations out, so you can listen to a paper on a walk and still follow the argument.

This guide covers getting the paper into the app, what the narration skips, how to change that, and what still doesn't work.

Get the paper into Narratorr

A paper is a PDF like any other, so any of these work:

1. In the Library, tap + and import the PDF from Files.
2. In Files, Mail or Safari, touch and hold the PDF, choose Share or Open With, and pick Narratorr.
3. If you are reading the paper in Narratorr's browser, choose Add PDF to Library from the browser menu.

The paper opens like a book. The support page has more on importing, and How to turn a PDF or EPUB into an audiobook on iPhone covers PDFs in general.

How a paper is read

Narratorr detects papers on its own; there is nothing to switch on. When it finds one, it reads it the way a person would:

• Columns in order. The left column, then the right, instead of jumping across the page line by line.
• Whole paragraphs. A paragraph that continues in the next column or on the next page is joined back together, so a sentence isn't cut in half by a page break.
• Title, then authors. The paper starts with what it is and who wrote it.
• No tables. Tables are left out of the narration; a grid of numbers read aloud is noise.

The layout setting is called Research Paper Layout. Open the paper, tap Aa on the playback bar to open Themes & Settings, and look under Customize.

On top of that, every PDF gets Clean PDF reading: running headers, footers and page numbers are skipped, and words split across lines with a hyphen are rejoined. If a document reads oddly because of it, turn it off with Clean Up PDF Text in reading settings.

What the narration skips

By default, a paper is read without:

• citations like [12] or (Smith et al., 2020)
• footnote markers
• the reference list at the end

Web addresses are read as "link" rather than letter by letter.

Nothing is deleted. Skipped text stays on the page, shown fainter, so you can still see what was cited when you look at the screen.

Choose what to skip

What narration leaves out is up to you, and the setting applies to everything you read, not only papers. Open Speech Settings (the waveform button at the left of the playback bar) and open Skip While Reading. You can choose to leave out:

• citations
• footnote markers
• links: read them, say "link", or leave them out
• text in brackets or parentheses
• references
• footnotes
• figure and table captions

Some combinations worth knowing:

• Listening for the argument. Skip citations, footnote markers, references and captions. The paper reads like an essay.
• Checking the sources. Turn citations back on when the "who said it" matters, for example in a literature review.
• Dense humanities papers. Footnotes often carry half the argument. Keep them, and skip only the markers.

Bracket and parenthesis skipping is the strongest of these. It also drops asides like "(see Figure 3)", which is usually what you want, but sometimes removes a definition. Try a page with it on before committing to a long paper.

Pick a voice

There are three ways to have the paper read, and you can switch at any time in Speech Settings:

• Offline AI. Natural AI voices that run on the iPhone itself. Free with no limit, no internet needed, and it keeps reading with the screen locked. It needs an iPhone or iPad with an A15 or M1 chip or newer (iPhone 13 and later). As of October 2026 it is one download of about 111 MB for all seven languages.
• Online AI. Natural voices that run on our servers. One free hour, then $2.99 a month or $19.99 a year. See pricing.
• Device voices. Your iPhone's built-in voices: free, unlimited and offline.

For a long paper, Offline AI is the usual choice: it doesn't use your free hour and works on a plane. Offline text to speech on iPhone explains the setup.

Follow along, or take it with you

If you read along while listening, the spoken word is highlighted. Themes & Settings also has reading aids under Customize, like a reading ruler that follows the line being read, with one or three lines and adjustable dimming.

To listen without the app open, export the paper. In the Library, touch and hold it, choose Export as Audio, and pick Audiobook (one M4B file with chapter markers) or Chapter Files, with Offline AI or a device voice. Online AI voices can't be exported. The files stay in Narratorr until you share them to Files, AirDrop or an audiobook app.

What doesn't work

Plainly, so you know before you import a stack of papers:

• Equations and tables. Tables are skipped, and a paper whose content lives in its equations is hard to follow by ear.
• Scanned papers. An old paper scanned as images has no text to read. On iOS 26, Narratorr reads scanned PDF pages with Apple's on-device text recognition; the result is only as good as the scan.
• Locked PDFs. A PDF protected with DRM can't be opened. Narratorr says so instead of showing blank pages.
• Unusual layouts. A poster, a slide deck saved as PDF or a page full of sidebars is not a paper, and may not come out in the order you expect.

Where to go next

The features in this guide arrived in version 1.11; the changelog lists everything in it. For anything not covered here, the support page answers the common questions, and Narratorr is on the App Store for iPhone and iPad.