Bring what you already have

Six formats in. One thing out.

You do not need to prepare anything. Drop in the file as it is — Unmixr reads the document's own structure, so chapters arrive as chapters instead of one four-hour blob.

PDF Papers, reports, scanned-to-text books
EPUB Ebooks, with their chapters intact
DOCX Drafts, memos, manuscripts
TXT Plain text of any length
MD Notes and technical documentation
HTML Saved articles and web pages

No file? Paste the text straight in and it starts reading immediately — no upload, no parsing step.

Read Aloud demos

Don't take our word for it. Press play.

Ten demos produced for this page — six kinds of document, each read with the register that document actually needs, then one identical paragraph read by four narrators so you can choose who to spend two hours with. Every script is printed beside the audio; headphones recommended.

  • Original generated audio
  • Scripts included
  • No sign-up to listen
Six kinds of document

A paper, a contract, a runbook. All read properly.

A narrator that reads a termination clause the way it reads a magazine feature is reading one of them wrong. Hear six document types given the register each actually needs — a finding that has to land separately from the method, figures that need a beat, a warning you cannot miss while your hands are busy.

6 document types

Can text-to-speech handle real documents? The question is really about register: a clause has to be unambiguous, a definition has to be slow enough to note down, and a procedure has to pause where you would act. All three are directable per document.

Milo Study narrator
PDF · 24 pages

Research paper

We tracked sleep onset in four hundred and twelve adults over eleven weeks. Participants who stopped screen exposure ninety minutes before bed fell asleep, on average, seventeen minutes faster; the effect held after controlling for caffeine, age, and exercise.

Unmixr AI Voice
0:20
Rafi Briefing narrator
DOCX · 40 pages

Business report

Revenue closed the quarter at nineteen point four million, up eleven percent year over year. Renewals carried it; new logos came in under plan for the third quarter running. That gap is the decision in front of this board.

Unmixr AI Voice
0:14
Nora Document narrator
PDF · 8 pages

Policy and contracts

Either party may terminate this agreement with thirty days' written notice. Termination does not release the client from fees already accrued, and any material shared under clause four remains confidential for two years after the agreement ends.

Unmixr AI Voice
0:20
Sia Study narrator
EPUB · 12 chapters

Textbook chapter

A cell membrane is not a wall. It is a gate that decides, molecule by molecule, what gets through. Hold on to that image, because almost everything in this chapter follows from it.

Unmixr AI Voice
0:16
Theo Long-read narrator
HTML · 2,400 words

Saved article

The reading list is not a reading list. It is a monument to the person you intended to become this year. Forty tabs, eleven saved newsletters, and one book you have opened to page nine on four separate occasions.

Unmixr AI Voice
0:14
Arne Procedure narrator
MD · 1,100 words

Technical docs

Drain the node before you restart it. Wait for active connections to reach zero — this usually takes under a minute. Do not skip this step on a primary; restarting a primary with live connections will drop every request in flight.

Unmixr AI Voice
0:18

Every one of these came from a file, not a script. Upload the PDF, DOCX, or EPUB you already have and press play.

Listen to your document
Pick who reads to you

Same paragraph. Four narrators. Two hours.

The choice that matters for a long listen is not the model — it is whose voice you want in your ears for the next two hours. The same paragraph, read four ways, so the comparison is like for like.

100+ languages available

Can I change the narrator partway through? Yes — switch voice in the player and the rest of the document re-voices as you listen. You do not restart, and you do not lose your place.

Milo Narrator
Books and long articles

Warm and low

He had assumed the hard part would be the climb. It was not. The hard part was the hour before dawn, sitting with a cold cup of coffee, deciding all over again that he was going to do it.

Unmixr AI Voice
0:14
Nora Narrator
Reports and policy

Clear and even

He had assumed the hard part would be the climb. It was not. The hard part was the hour before dawn, sitting with a cold cup of coffee, deciding all over again that he was going to do it.

Unmixr AI Voice
0:14
Sia Narrator
Study material

Bright and patient

He had assumed the hard part would be the climb. It was not. The hard part was the hour before dawn, sitting with a cold cup of coffee, deciding all over again that he was going to do it.

Unmixr AI Voice
0:14
Arne Narrator
Technical documents

Firm and precise

He had assumed the hard part would be the climb. It was not. The hard part was the hour before dawn, sitting with a cold cup of coffee, deciding all over again that he was going to do it.

Unmixr AI Voice
0:14

Four are playable here. The full library covers 100+ languages, and you can clone an approved voice where you have permission.

Try a narrator
The reader

Built for two-hour sessions, not two minutes.

Chapters down the side, the paragraph you are on lit up in the page, and a player that keeps your place. Click any paragraph to jump the audio there.

Chapters

Follow along, or don't

The text stays on screen with the current paragraph highlighted, and Follow mode scrolls to keep it in view. Turn it off and it becomes a plain audio player.

Resumes at 1.25x Chapter 3 Follow on

It remembers where you stopped

Position, speed, theme, and font size are kept per document. Come back next week and the report picks up mid-paragraph at the pace you had already settled on.

How it works

Upload. Pick a voice. Listen.

Four steps, and the only one that takes effort is picking the file.

1

Drop in the document

PDF, EPUB, DOCX, TXT, Markdown, or HTML — or paste text if you have no file. Unmixr reads the document's own structure, so chapters arrive as chapters and the cover page is skipped.

2

Choose who reads it

Audition narrators on the document in front of you rather than a stock demo line. You are about to spend hours with this voice; pick it on the real thing.

3

Listen and follow along

The text sits on screen with the current paragraph highlighted. Click any paragraph to jump there, search the whole document, and set the speed anywhere from half to double.

4

Take it with you

Export MP3 per chapter or for the whole document for the commute, the gym, or a flight. Close the tab and your position is still there tomorrow.

What makes it a reader

A document, not a wall of audio.

Every one of these exists because a long listen breaks down without it. Structure, position, and voice choice are the difference between finishing a book and abandoning it in chapter two.

Chapters survive the upload

EPUB spines, DOCX heading levels, and PDF outlines are read as structure, not scraped as one wall of text. You get to jump to chapter nine instead of scrubbing a four-hour file.

It remembers where you stopped

Position, speed, theme, and font size are kept per document. Come back a week later and the report picks up mid-paragraph at the pace you had already settled on.

Change the narrator mid-listen

Switch voice in the player and the rest of the document re-voices as you go. No restart, no re-upload, no losing your place forty minutes in.

Search the whole document

Ctrl or Cmd+K opens search across every chapter. Pick a result and the reader moves there — useful when you half-remember a line and need to hear its paragraph again.

Speed that actually holds up

0.5× to 2×, changeable while it plays. Slow down for the methodology, run at 1.6× through the parts you already know.

Built for long sessions

Light, sepia, and dark reading themes with adjustable type, plus keyboard control and lock-screen playback. It is a reader, not a media player with text bolted on.

Honest comparison

My phone already reads text aloud.

It does, and for a quick page that is genuinely enough. The gap shows up at hour two.

A built-in screen reader

  • Reads what is on screen, in order, ignoring chapters
  • One system voice you cannot change
  • Loses your place when you close the app
  • Nothing to download and take offline

Unmixr Read Aloud

  • Reads the document — chapters, sections, skipped cover pages
  • Choose your narrator, and change it mid-listen
  • Keeps position, speed, and theme per document
  • Export MP3 per chapter for the commute
Who uses it

For researchers, students, analysts, and anyone with a reading backlog.

The common thread is not the document type. It is having more to read than there are hours to sit still and read it.

Papers and research

Get through the literature review on the walk instead of letting twenty PDFs age in a folder.

Reports and board packs

Hear the forty-page pack before the meeting rather than skimming the summary in the lift.

Books and long reads

Ebooks you own, chapter by chapter, with your place kept between sessions.

Publishing a book instead? →

Study material

Textbook chapters and lecture notes read at a pace you can take notes against.

Narrating a course instead? →

Contracts and policy

The documents nobody reads carefully because reading them carefully is unpleasant. Have them read to you properly.

Accessible reading

A genuine alternative for dyslexia, low vision, screen fatigue, or anyone who simply retains more by ear.

100+ languages to listen in
6 file formats accepted, plus pasted text
0.5–2× reading speed, adjustable mid-listen
500M+ minutes generated on Unmixr
FAQ

Read Aloud: your questions, answered

Straight answers for people with more to read than time to read it.

What is Read Aloud?

Read Aloud turns a document you already have into something you can listen to. Upload a PDF, DOCX, or EPUB — or paste text — and Unmixr reads its structure, splits it into chapters and paragraphs, and narrates it with an AI voice you choose. It is built for the reading you keep meaning to get to, not for publishing audio to an audience.

Which file formats can I upload?

PDF, DOCX, EPUB, TXT, Markdown, and HTML. You can also type or paste text directly if you have no file. EPUB and DOCX carry their own chapter structure, so those come through with chapters intact; PDFs use the document outline where one exists and fall back to a clean text read where it does not.

Does it keep the chapters, or is it one long audio file?

It keeps the structure. Chapters become sections and paragraphs become individual blocks, so you can jump to chapter nine, replay one paragraph, or pick up exactly where you stopped. It is also why you can regenerate a single paragraph in a different voice without re-rendering the whole book.

Can I change the narrator partway through?

Yes. Switch voice in the player and the rest of the document re-voices as you listen — you keep your position and your place in the text. That is the point of the four-narrator demo on this page: the paragraph is identical so you can compare like for like before committing to a long listen.

Can I control the reading speed?

Yes, from 0.5× to 2×, adjustable while it plays. Speed, position, theme, and font size are remembered per document, so returning to a long report picks up where you left it at the pace you had set.

Can I follow along with the text while it reads?

Yes. The document is displayed as a book-style page with the paragraph being read highlighted, and Follow mode auto-scrolls to keep it in view. Click any paragraph to jump the audio there, and use Ctrl/Cmd+K to search the whole document and jump to a result.

Can I download the audio?

Yes — export MP3 per chapter or for the whole document, which covers offline listening on a commute or a run. The exports come from the same narration you approved in the player.

How is this different from my phone's built-in screen reader?

A screen reader reads what is on screen, in order, in one system voice. Read Aloud reads the document — it understands chapters, skips the cover page, keeps your position across sessions, lets you choose and change the narrator, and gives you a downloadable file. It is the difference between accessibility playback and something you would choose to listen to for two hours.

What languages does it support?

Narration is available in 100+ languages, and you can clone an approved voice in 80+ where you have permission. Pick the voice that matches the language your document is written in.

Is there a limit on document length?

Long documents are expected — books and reports are the normal case, and the reader handles multi-hundred-chapter structures. What you generate draws on your plan's quota, and Unmixr shows an estimate before it starts so there are no surprises. You can start free without a card.

Your reading list is not going to read itself

Upload one document, pick a narrator, and listen to the first chapter free. No card, no install.

Listen to a Document Free
Audiobook Creator Course Narration E-Learning Voiceover AI Text to Speech Voice Cloning Speech to Text Document Translation Pricing

Still have a question?

If you have a specific use case or need help configuring a custom workflow, our team is always here for you. We would love to listen to your needs!

support@unmixr.com Join the Discord Community GET SUPPORT