Turn a PDF into an audio show
Upload any PDF to Lissin and get a narrated show with a transcript. Scanned PDFs go through OCR automatically. What to ask for so a 90-page report becomes listenable.
The short answer
Upload the PDF and describe the show you want from it. Lissin pulls the text layer directly from a born-digital PDF; if there is barely any text — a scan, a photographed document — it escalates to vision-based OCR automatically. You do not choose between the two.
How to do it
Upload the file as it is
No pre-processing, no splitting, no converting to text first. A born-digital PDF is parsed straight from its text layer, which is fast and costs nothing. A scanned or image-only PDF falls through to an OCR pass that reads the pages as images, including handwriting and figures with labels.
Tell it how much of the document matters
A ninety-page report is not a ninety-minute episode. ‘Fifteen minutes on the findings and what they mean for a UK reader, skip the methodology’ is a real instruction and produces a real show. Without one, the script tries to be faithful to the whole document and the result is flat.
Ask for a format that suits the material
Dense argument benefits from two voices disagreeing; a set of results benefits from a single narrator walking through them in order. Both are available, and choosing deliberately is the difference between an episode you finish and one you do not.
At a glance
- Born-digital PDFs
- Text layer read natively
- Scanned PDFs
- Automatic vision OCR fallback
- Other formats
- DOC, DOCX, PPTX, XLSX, HTML
- Transcript
- Included, synchronised to the audio
Hear what comes out
Shows already published on Lissin, free to play in the browser.
Questions
- Does a scanned PDF work?
- Yes. When the native parse yields almost no text, Lissin sends the pages to a vision model instead, which handles scans, photographs of pages and multilingual documents. You do not have to flag it.
- Is this just text-to-speech on the PDF?
- No, and that is the main difference from a read-aloud app. Lissin writes a script from the document — restructuring, explaining and cutting — and then narrates the script. If you want the document read verbatim, ask for that explicitly.
- How long can the PDF be?
- Long documents are handled, but the useful limit is editorial rather than technical: past roughly a hundred pages you should be asking for a specific slice rather than the whole thing, or the episode has to skim to fit.
What this does not do
A PDF that is mostly tables or charts loses most of its meaning in audio, whichever extraction path runs. Password-protected files are not opened. And the script is an interpretation of the document, not a substitute for reading it when the exact wording matters — a contract, a filing, a spec.
Try it on something you already have
Upload a PDF, a deck or a document and describe the show you want from it. Or press play on anything in the library first — no account, no app.
Last reviewed 9 September 2026 by Lissin Editorial.