Text-based PDFs: excellent
Anything exported from Word, Google Docs or a publishing tool has a real text layer. These convert cleanly and are the common case.
Use case
Turn a PDF into audio without it sounding like a robot reading a fax
PDFs are the worst format to read on a phone and the most common format for things you have to read anyway. Converting one to audio is straightforward. Getting audio that is pleasant to listen to depends on what kind of PDF you are holding, and it is worth knowing the difference before you spend credits.
How it works
- Upload the PDF. PDF is a directly supported input, alongside Word, links and pasted text.
- Choose full narration or a spoken summary. Full narration for material you must actually know; a summary when you are deciding whether it matters.
- Wait a few minutes. Most documents finish in under five.
- Listen, download, keep. It stays in your library, groupable into playlists, resumable.
Not all PDFs convert equally well
This is the part most tools do not tell you. Check which kind you have before converting a 300-page file.
Scanned PDFs: check first
A photograph of a page has no text layer. If you cannot select the text with your cursor, no reader can find words to read either.
Heavy layout: expect roughness
Multi-column journals, dense tables, footnotes and figure captions were designed for eyes. Read in sequence they can interrupt the flow.
Formula-heavy pages: use a summary
Mathematics does not narrate well in any tool. A spoken summary of the argument is usually more useful than a literal reading.
Getting a clean result
- Try selecting text in your PDF viewer first. If you cannot highlight it, it is a scan and needs OCR before conversion.
- On a long document, convert a summary first. It costs far fewer credits and tells you whether the full narration is worth it.
- For journal articles, consider converting only the sections you need rather than the whole file including references.
- Credits are minutes, so a 40-page report costs what 40 pages takes to speak. Budget accordingly.
What a PDF will not give you
- A scanned page with no text layer cannot be read at all. This is the single most common failure - run OCR first.
- Tables, equations and figure captions carry meaning visually, and none of that survives being read in sequence.
- Multi-column journal layouts can interleave badly, so the reading order is occasionally wrong.
- Page headers, footers and reference lists are text too, so a literal narration includes them unless you trim first.
- No two-host discussion of the document - that format is marked coming soon.
Frequently asked questions
Will it read a scanned PDF?
Only if the scan has been through OCR and has a real text layer. Test it by trying to select the text with your cursor - if you cannot, neither can the converter.
How much does a long PDF cost?
One credit per minute of finished audio, so cost tracks length - a 45-minute narration is 45 credits. Budget by listening time rather than by file count, and note that a failed generation is refunded automatically.
Can I get an MP3 out?
Yes, Loudly produces downloadable audio.
What about the references section?
A literal reading includes whatever text is in the document. For papers, a spoken summary or converting only the sections you care about usually gives a better listen.
Start with the PDF you have been avoiding
Upload it and listen to the first few minutes.
No commitment - credit packs, cancel anytime. · 1 credit = 1 minute of finished audio