Use case
PDF to podcast: turn a document into an episode
A quarterly report, a thesis, a policy brief, a manual. Upload it and PodThis reads it, keeps the facts, adds context, and produces a voiced episode with chapters and a transcript in minutes.
annual-report-2025.pdf — 48 pages, condensed to notes the writer can work from
- Upload the document
PDF, Word (.docx), .txt or .md, up to 20 MB and 200 pages. Tick the box confirming you have the right to use it. The file is read in memory and never stored.
- Check the text
The extracted text appears in the material box. Long documents are condensed into faithful notes that keep every figure, name and claim. Trim what you don’t want.
- Generate
Pick a style and length. The pipeline researches around your document, writes an audio-first script that cites it, and voices the episode with a full cast.
What happens to the document
The File tab in the studio accepts PDF, Word, plain text and Markdown. The text layer is extracted on our server and, when it is longer than the studio’s 20,000-character material limit, condensed by Gemini into working notes that keep the document’s own order, figures, dates, names and claims. Scanned PDFs without a text layer are refused with a clear message, not silently turned into a guess.
The file itself is not kept. What stays with the episode is the condensed text, the file name and a checksum, so the episode page can say which document it was made from. Uploads are capped at 20 MB and 200 pages, and every import requires you to confirm you have the right to use the material.
What the episode sounds like
The script treats your document as the primary source: the hosts attribute claims to it, stay faithful to its numbers and do not contradict it with outside guesses. Around it, the research stage adds the context a listener needs. Every style works, from a two-voice explainer to a narrated documentary, and the episode page carries the full transcript and a “Sources & further reading” block that names the document first.
Why listen to a document at all
Dense documents get read at a desk, in one sitting, or not at all. As audio content they fit the walk, the commute and the gym. An AI podcast that explains the document in order is also easier to follow than a synthetic voice reading every line and footnote, and its chapters keep the document’s structure, so you can go back to the section you need.
Unlike a plain reader, the podcast generator does not stop at the page. It researches the context a listener needs: what the terms mean, what happened before, why the numbers matter. Your document stays the primary source and the context sits around it. That works for a wide range of material, from a ten-page brief to a 200-page report condensed into notes.
Good documents for this
Reports and white papers, lecture notes and course readers, meeting minutes you need to absorb on a commute, grant applications, product documentation. Anything with dense text and a structure. Slide decks with little text and scanned books without OCR are the two cases that do not work. For which files work, how to prepare them and which style to pick, read the step-by-step guide to turning a PDF into a podcast.
Episodes made from documents
“amundsen-polar-plateau-december-1911.pdf”
Make yours
Every account starts with 5 credits and gets 3 more each month. A short episode costs one. Every feature is on every plan.