I have a folder on my laptop called to-read that, at any moment, holds about thirty PDFs I fully intended to read and never did — a lease, two course readers, a 60-page whitepaper a coworker swears is "actually good." At some point I stopped trying to read the backlog and started trying to listen to it instead, on walks and on the train where my hands are busy holding a pole.
What I learned is that "play this PDF out loud" sounds like one task and is actually two, and most tutorials skip the part that decides which one you're in. So here's the practical version: how to tell what kind of PDF you have, every free way I've found to listen to it, and the honest "skip this one" notes for each. None of it requires paying anyone — and CastReader, which I'll mention where it fits, is free across the board.
The one test that decides everything: text PDF or scanned PDF
Before you pick a tool, do this. Open the PDF and try to drag-select a single sentence with your cursor. Better yet, press Ctrl/Cmd+F and search for a word you can clearly see on the page.
- If the word highlights and search finds it, you have a text PDF. The letters are real, selectable characters. Almost anything exported from Word, Google Docs, LaTeX, or a journal site is text-based. This is the easy case, and every method below will work.
- If nothing selects, the whole page highlights as one block, or search finds zero results, you have a scanned PDF. It's really just a photo of each page — an image with no characters underneath. Old books, library scans, signed contracts, and a huge share of "free PDF" downloads are like this.
This matters because most read-aloud tools can only speak text that already exists. Point a basic reader at a scanned PDF and you get dead silence — there's literally nothing for it to read. To listen to a scan, the text has to be recovered first with OCR (optical character recognition), covered below. Do the 10-second select test now; it'll save you from blaming a tool that was never going to work.
Method 1: Your browser already reads text PDFs (kind of)
If you just want a text PDF spoken right now, the fastest free option is the browser you already have.
Microsoft Edge is the standout. Open any text PDF in Edge, click Read aloud in the PDF toolbar (the "A" with sound waves, or right-click → Read aloud), and it starts narrating. The voices are surprisingly decent — Edge ships with Microsoft's neural voices — and you get a floating bar with play/pause and a speed slider. For a quick, no-install read of a normal PDF, this is genuinely good.
Chrome is weaker: no real built-in "read this PDF" button, just its general screen-reader support, which is clumsy for documents. For PDFs specifically, Edge wins among browser built-ins.
Don't use it when: the file is scanned (Edge will read nothing — no OCR), or the PDF is two-column. Browser readers tend to read straight across both columns instead of down one, so sentences interleave into nonsense. The moment a passage stops making sense, that's almost always why. Also, Edge keeps you parked at the desk — there's no "keep listening on my phone."
Method 2: Adobe Reader and your OS already have read-aloud built in
Two more free options are probably already on your machine.
The free Adobe Acrobat Reader has a feature buried under View → Read Out Loud → Activate Read Out Loud, then Read This Page (or the whole document). It works, but it feels like 2008: it uses your OS's default voices, which can sound robotic, and the controls are clunky menu-diving. It earns its place on plain single-column PDFs and locked-down work computers where you can't install anything new.
Your operating system has a reader too. On macOS, turn on Spoken Content in System Settings → Accessibility, set a shortcut, select text, and hit it — the newer "Siri" voices are genuinely natural. Windows has Narrator (Win+Ctrl+Enter), though it's built for full screen-reading and is heavy-handed for documents. The catch with the OS route: it reads whatever you select, so you're constantly re-selecting text as you move down a PDF, headers and page numbers included.
Don't use these when: you want continuous, natural narration through a long document (the re-selecting and dated voices defeat the purpose), or the PDF is scanned — like the browser, both read the text layer and go silent on pure images.
Method 3: A dedicated reader (handles scanned PDFs and follows you to your phone)
The methods above share two blind spots: they choke on scanned PDFs, and they leave you tethered to one screen. A dedicated text-to-speech reader closes both gaps, and it's where I've ended up for anything longer than a page or two.
This is what CastReader is built for, and it's free. The two things that actually matter:
- It OCRs scanned PDFs. Instead of giving up when there's no text layer, it reads the page the way your eyes do — by looking at the rendered pixels and decoding the actual letters — so a photographed book or signed contract becomes audio instead of silence. (Same trick that lets it read Amazon's scrambled Kindle text, where the underlying character codes are deliberately obfuscated.)
- It moves between your devices. Start a document on the Chrome/Edge extension or the Mac app at your desk, then keep listening on the iOS or Android app on your commute — see send to phone. That's the difference between "read me this paragraph" and "turn this PDF into a podcast I finish on a walk."
The basic flow for a text PDF: open the document, let it pull the text, pick a natural voice, press play, set speed to taste (1.0–1.25x for dense material, 1.5x+ once you're warmed up). For the dedicated PDF to audiobook experience that keeps your place, that's the path. And because the same engine reads any text on the web, the moment you're done with PDFs it also handles your Google Docs, Notion, and long Claude and ChatGPT threads.
Don't use it when: honestly, the only weak case is a single short paragraph you need read right now and you have no reader installed — for that one-off, Edge's built-in or your OS reader is faster than installing anything.
Getting clean audio from scanned PDFs
If your test came back "scanned," here's how to make OCR actually pleasant rather than maddening.
Scan quality is everything. A crisp 300-DPI scan OCRs almost perfectly. A crooked phone photo with shadows and highlighter marks will produce errors — OCR famously reads "rn" as "m," mangles accented names, and trips on handwritten margin notes. If you control the scan, scan straight, in good light, at high resolution. Cleaning the source beats fighting the output every time.
A few realities to plan around:
- Headers, footers, and page numbers get read aloud. "Chapter 4. Thermodynamics. 87." sprinkled through your audio is normal for raw OCR — mildly annoying, but your brain learns to tune it out.
- Tables and figures don't narrate. A grid of numbers read linearly is meaningless; when you hit one, skip ahead. Charts are the one thing audio genuinely can't replace, so keep the document open for those and listen for the prose around them.
Academic PDFs, equations, and one format tip
Research PDFs are their own genre, and a few things matter for audio. Two-column layouts are the number one reason a paper reads in the wrong order — basic extraction sometimes reads across both columns. If you read papers often, prefer a single-column source; many arXiv papers now offer an HTML version that's one clean column and narrates far more reliably than the PDF. Equations are a lost cause by ear ("x subscript i equals sum…"), so put the prose on audio and look at the math when narration reaches it. And stop before the reference list — there's no listening value in forty "Smith, J., 2021, Proceedings of…" entries.
The bigger tip: if you can choose the format, choose EPUB. A PDF is a fixed picture of a page; an EPUB is reflowable text with real chapter structure — no column traps, no page-number litter — and the EPUB to audio reader flow is far smoother than any PDF. This is also why listening is such a quiet win for students and for readers with dyslexia or ADHD: hearing the text alongside seeing it cuts the effort of getting through it.
What about the paid readers?
It's fair to ask why not just buy a polished app. Speechify and NaturalReader both read PDFs well and have nice voices — but the genuinely good voices and unlimited listening sit behind subscriptions. Speechify's premium runs about $139/year; NaturalReader's paid plans are in a similar range. If you only need to clear a backlog of documents now and then, paying a yearly subscription for it is a hard sell. If you're weighing them, I wrote honest side-by-sides at CastReader vs Speechify and CastReader vs NaturalReader.
CastReader's whole pitch is that listening to a PDF should be easy: the free tier reads any PDF aloud in a natural voice with full speed control and no signup, and CastReader Pro adds premium ultra-realistic voices, more listening hours, and AI document analysis if you want them. The extension is in the Chrome Web Store, the apps are on the App Store and Google Play, and there's a Mac app for when you'd rather not keep a browser tab open all day.
FAQ
Can I listen to a scanned PDF out loud for free?
Yes, but only with a tool that runs OCR first, because a scanned PDF is just images with no text underneath. Browser, OS, and Adobe readers will go silent on a scan. A reader with built-in OCR (like CastReader) reads the visible letters from the page pixels and speaks them. Clean, high-resolution scans work best; blurry phone photos introduce errors.
Why is my PDF being read in the wrong order?
Almost always a two-column layout. Text extraction sometimes reads straight across both columns instead of down one, so sentences interleave into nonsense. It's most common in academic papers and textbooks. Skip to the next clean paragraph, or use a single-column source — like the HTML version of an arXiv paper — when one exists.
What's the easiest free way to read a normal text PDF aloud?
For a quick one-off, open it in Microsoft Edge and click Read Aloud — no install, decent neural voices. For anything longer, or to keep listening on your phone, a dedicated reader is worth the one-time setup because it handles continuous playback and scanned files.
Will any of this read my equations and tables?
Not usefully. Equations narrated by ear ("x subscript i…") are nearly impossible to follow, and tables read linearly are meaningless. Put the prose on audio and look at the math, tables, and figures with your eyes when narration reaches them.
Is CastReader free?
Yes — CastReader is free to use across the Chrome/Edge extension, the Mac app, and the iOS and Android apps: read your PDFs, papers, and ebooks aloud in a natural voice with no signup. There's also an optional CastReader Pro plan that adds premium ultra-realistic voices, more listening hours, and AI document analysis if you want them.
The short version: run the 10-second select test first. If it's a text PDF, Edge or your OS reader will get you going in seconds; if it's scanned or you want to listen across a walk and a commute without re-selecting anything, reach for a dedicated reader that can OCR and follow you to your phone. Either way, that to-read folder stops being a guilt pile and starts being a queue you can actually clear.
Got a PDF that refuses to cooperate? Email us at support@castreader.ai — we read every message.