Extract Text from PowerPoint

1. 1

   Add a presentation

   Drag in a .pptx, .pptm, .ppsx, .potx, .odp or .ppt file, or browse for one. It is read on your device and never uploaded.

2. 2

   Choose what to include

   Turn speaker notes, tables, chart labels, SmartArt, comments and alt text on or off. Set a slide range if you only want part of the deck.

3. 3

   Pick an output format

   TXT for reading, Markdown for pasting into a document, HTML for the web, JSON for scripts, or CSV with one row per shape for translation.

4. 4

   Copy or download

   Copy the result to the clipboard, or download it. With Pro you can queue up to 50 presentations and take them all as one archive.

It reads a presentation the way PowerPoint reads it, and gives you back everything that was written in it - not just the words in the obvious text boxes.

• Every slide, in presentation order: taken from the slide list the file stores, so a reordered deck comes out in the order you present it.
• Speaker notes on the right slide: matched through the file's own relationships rather than by counting, which is what breaks when one slide has no notes.
• Structure that survives as text: paragraphs, line breaks, bullet levels, numbered lists, and tables as rows and columns.
• Text most extractors miss: chart titles, series names and axis labels, SmartArt, comments with their authors, and picture alt text.
• Five output formats: TXT, Markdown, HTML, JSON and a per-shape CSV built for translation work.
• Three file families: PPTX and its variants, OpenDocument .odp, and PowerPoint 97-2003 .ppt.
• Nothing uploaded: the presentation is opened on your device and never leaves it.

The list is closed and it is enforced when you pick the file, not after a spinner. Anything not on it is refused with a message that names what the file actually is and links to the tool that handles it.

• PPTX, PPTM, PPSX, PPSM, POTX, POTM: PowerPoint 2007 and later. The same container underneath, so the extension makes no difference to what comes out.
• ODP and OTP: OpenDocument presentations from LibreOffice, OpenOffice and Google Slides exports, including their charts.
• PPT, PPS, POT: PowerPoint 97-2003. Read directly, with slide order and notes resolved properly - most tools ask you to convert these first.

Sent somewhere better

• PDF: the PDF to Text extractor reads those.
• Word documents and RTF: Word to Text handles .docx, .odt, .rtf and .doc.
• Spreadsheets: Excel to CSV covers .xlsx, .xls and .ods.
• Images and screenshots: Image to Text recognises words in a picture; this tool only reads real text objects.
• Password-protected files: open the file in PowerPoint, remove the password and save a copy.

Inside a .pptx the slides are separate files called slide1.xml, slide2.xml and so on, and it is tempting to read them in that order. The numbers are assigned when a slide is created and never change again. Move a slide to the front of your deck and its file keeps the number it has always had.

The real order lives in a list the presentation keeps for exactly this purpose, and that list is what this tool follows. The difference only shows up on decks that have been edited, which is to say almost all of them - and it shows up as output that is quietly, plausibly wrong rather than as an error. Nothing about a deck extracted in the wrong order looks broken.

The same principle governs speaker notes. A notes page is joined to its slide by a named relationship inside the file, and notes pages are numbered independently of slides. Line the two lists up by position and a deck whose first slide has no notes gets every note attached one slide too early, with the last slide's notes lost. This tool follows the relationship.

A slide is not one box of words. Switch each of these on or off in the panel; anything the presentation does not contain is greyed out rather than left there doing nothing.

• Shapes and text boxes: always included, with the title placeholder identified so Markdown and HTML can use it as a heading.
• Tables: kept as rows and columns, as tab-separated lines, as one cell per line, or left out.
• Speaker notes: off by default, and placed either with each slide or gathered at the end.
• Charts: the title, the series names and the category labels, read from the chart's own stored copy of them - the labels a viewer actually sees.
• SmartArt: the text of each node, without the duplicates that the graphic keeps for layout.
• Comments: both the classic and the modern threaded format, with author names and replies. Off by default, because a comment is about the deck rather than part of it.
• Picture alt text: what a screen reader would announce. Useful for accessibility audits and for translating a deck completely.
• Footers, dates and slide numbers: off by default, along with generated field text. Leaving them on is how an extractor reports a slide with no notes as having notes that read “1”.

The same extraction, arranged five ways. Switching between them is instant - nothing is read again.

• Text: slide headers you can change, bullets, tables as tab-separated rows. The format to paste anywhere.
• Markdown: a heading per slide, real bullet lists with their nesting, real tables, notes as block quotes, and bold and italic carried through from PPTX and ODP.
• HTML: one section per slide with headings, lists and tables, with bold and italic as strong and em. From PPTX and ODP; the 1997 binary format keeps character formatting somewhere this reader does not go.
• JSON: every block with its slide number, kind, shape name and nesting level, for anything that reads a deck programmatically.
• CSV: one row per shape - slide, source, shape name, level, row, column, text - written with CRLF and quoting so Excel opens it correctly. This is the shape a translation memory or a localization tool expects.

Sending a whole presentation to a translator means sending the design with it. Exporting the CSV instead gives one row per piece of text, each still labelled with the slide it came from and what kind of element it was, which is what makes the translated file straightforward to check and to put back.

For that work, turn speaker notes on - narration is usually where the real wording lives - and turn on picture alt text so nothing accessible is left in the original language. Chart labels and SmartArt are on already. A slide range keeps a long deck to the section being worked on.

A UTF-8 file with no byte-order mark opened in Windows Notepad, or imported into Excel, is read as the system code page - so every accent and every non-Latin script arrives as mojibake, and by then the file is written and the reader cannot fix it. UTF-8 with a BOM avoids it. UTF-16 is there for older Windows tools that still expect it. CRLF line endings are the Windows convention; LF is what everything else uses. The CSV export always uses CRLF regardless, because that is the one format whose main consumer is Excel on Windows.

Most of a presentation's size is images. A reader that unpacks the whole file to get at the words spends nearly all of its time on bytes it will never look at - measured on a 200-slide, 16 MB deck, 88 percent of the old reader's time went on exactly that.

This one reads the archive's index, opens only the parts that carry text, and never touches the images at all. It also runs off the main thread, so the page keeps responding while a long deck is read, and every setting re-renders from what is already in memory instead of reading the file again.

The presentation is opened by the page you are looking at. It is not uploaded, not queued on a server, and not stored - there is nothing to delete afterwards. Once the page has loaded the tool keeps working with the network off, which is the simplest way to be sure. That makes it usable for material you could not put through an online converter at all: unreleased results, client work, anything under NDA.

Only the parts of a presentation that carry text are ever opened, so a large deck costs about as much as a small one.

• Word Count in PowerPoint
• Split PowerPoint
• Merge PowerPoint
• Compress PowerPoint
• Word to Text
• PDF to Text
• Excel to CSV
• Image to Text
• PowerPoint Tools