What Read Aloud does well
There is nothing to install and nothing to pay. Open a page or a PDF in Edge, start Read Aloud, and the text is read with each word highlighted, at the speed and in the voice you choose. For reading support, tired eyes or getting through a long article while you cook, it is excellent.
If all you want is to hear what is on your screen right now, keep using it. A separate tool only earns its place when you need the audio to exist after you close the tab.
Edge Read Aloud vs AI Audiobooks
Edge adds features over time, so its column is a general description; check Microsoft’s current documentation for details.
| AI Audiobooks | Edge Read Aloud | |
|---|---|---|
| Best for | Making audiobook files to keep or share | Listening to pages and PDFs as you read |
| Cost | Free import and casting; credits to generate | Free in the Edge browser |
| Saving audio | MP3 per chapter, ZIP or M4B | Not designed for it |
| Voices | A narrator and a separate voice per character | The voice you select, for all text |
| Chapters | From the EPUB contents or document headings | Not applicable to live reading |
| Inputs | Uploads of EPUB, Word, PDF, text and Markdown files | Web pages and PDFs opened in Edge |
| Control | Bracket directions and sound tags per line | Voice and speed |
Many people use both: Read Aloud for everyday reading in the browser, and the studio for the books they want to keep, share or publish.
Why recording your speakers is a poor workaround
Capturing Read Aloud with a sound recorder does produce a file, but it takes as long as the text itself, often nine hours or more for a novel. It catches notification pings, loses quality, and leaves you to cut chapters and add titles by hand. Generating the audio directly is faster and gives you clean chapter files.
From a web page or PDF to an audio file you keep
PDFs. Upload a text-based PDF and it is reflowed into chapters, with page numbers and running headers dropped so they are not read aloud. The PDF to MP3 page shows what the output looks like. Scanned PDFs made of page images are not supported, because there is no OCR.
Web pages. There is no URL import, so copy the text and paste it into the studio. Headings in the pasted text become chapter breaks, and quoted dialogue is cast like any book.
Ebooks and manuscripts. DRM-free EPUB files keep their table of contents as chapters, and Word files split on Heading 1 and Heading 2. Front matter such as the copyright page is set aside.
The files. Download one MP3 per chapter, a ZIP of everything with a chapters index, or one M4B with named chapter markers, title, author and cover. M4B vs MP3 helps you choose.
Directing the voices instead of only choosing them
Read Aloud gives you two dials: which voice, and how fast. An audiobook needs more, because the same voice may have to whisper in one scene and shout in the next. Here a short note in square brackets at the start of a line, such as [whisper], [furious] or [like a late-night radio host], changes how that line is delivered, and inline tags such as [sigh] add a sound exactly where you place them.
You also choose the model. Standard voices (Gemini Flash TTS) cost about $9 to $12 per finished hour, Lite is cheaper for drafts and long nonfiction, and Premium and Expressive use ElevenLabs voices, with Expressive the most emotional. Narration runs at a natural 150 to 160 words per minute, and listeners can still speed it up in their player.
Keep using Read Aloud for these
A browser feature is the right tool for plenty of everyday listening.
- News, blog posts and email you want to hear once
- Work documents and reports you read at your desk
- Following along with highlighted words for reading support
- Quick checks of how a paragraph of your own writing sounds
Turning what you hear in Edge into files
- 1
Get the source text
Use the PDF or ebook file itself, or copy the text from the web page.
- 2
Upload or paste it
Files open in your browser and the chapters appear before you spend anything.
- 3
Choose voices
Pick a narrator you would happily hear for hours and check the detected characters.
- 4
Generate and download
Generate a short sample using gift credits, then the rest, and save MP3 files or one M4B.

