How to convert PowerPoint to video with AI voiceover
By Omprakash Sah Kanu · June 22, 2026 · Updated August 9, 2026 · 13 min read

There are two practical ways to turn a PowerPoint into a narrated video: record and export it inside PowerPoint, or upload the deck to a tool that generates editable narration from each slide. PowerPoint can record voice, timings, pointer movements, and video, then export an MP4. With ScreenDub, you upload a PPT, PPTX, or PDF, review the narration it drafts from the slide content, choose a voice, and export a narrated video plus a written handout.
Choose a method
| Method | Best when | Main work | Output |
|---|---|---|---|
| PowerPoint recording | You want to narrate and control the presentation yourself | Record voice, timings, and optional camera or pointer gestures, then export | PowerPoint presentation plus MP4 or WMV video |
| AI narration from a deck | You want a draft script and do not want to record a microphone track | Upload, review each line, choose a voice, and export | Narrated MP4 plus a written slide-and-script handout |
| Manual editor | You need complex animation, compositing, or multi-track editing | Export slides, record or generate audio, then synchronize everything | A manually edited video with the editor’s supported formats |
If you need to preserve PowerPoint animations, embedded media, or a live presenter, test the native PowerPoint workflow first. If your goal is a self-running deck with generated narration and a reusable handout, a slide-to-video workflow is usually simpler.
Prepare the deck for narration
- Give each slide one job. A slide that tries to explain several unrelated ideas will produce a long, difficult narration.
- Use readable text. Viewers may watch the video in a small player, so avoid dense paragraphs and tiny labels.
- Make the takeaway visible. The headline should tell the viewer what the slide means, not only name the topic.
- Check figures and names. A generated script can read the wrong number or pronounce a product term incorrectly. Treat the first draft as an editing starting point.
- Export Google Slides or Keynote first. ScreenDub accepts PDF, PPT, and PPTX; export other presentation formats to one of those before uploading.
Method 1: export a narrated video from PowerPoint
PowerPoint’s native workflow lets you record narration and timings, then export the presentation to a video file. The exact menu names vary by PowerPoint version, but Microsoft’s current instructions use File → Export → Create a Video after the recording is complete. The export can use recorded timings and narrations or a fixed number of seconds per slide. See Microsoft’s official instructions for version-specific details.
- Save the current presentation as a PPTX.
- Record the slide show or add narration and timings to individual slides.
- Review the recording, especially slide transitions and pronunciation.
- Choose the export quality and whether to use recorded timings.
- Export as MPEG-4 if you want broad video-player compatibility.
Method 2: use ScreenDub for AI narration
- Export or save the deck. Upload a PPT, PPTX, or PDF. A Google Slides or Keynote deck should be exported first.
- Upload the presentation. Each slide becomes a video segment. The deck order becomes the video order.
- Review the draft. ScreenDub reads the slide content and creates an editable narration line for each segment.
- Rewrite instead of repeating. Keep the headline and key figures visible on screen, then use the voice to explain context, implications, and transitions.
- Choose a voice and export. Preview the voice against the deck, then export the narrated MP4 and written handout.
ScreenDub accepts files up to 650 MB and up to 100 slides per deck. The per-video length cap depends on the plan: 6 minutes on Free, 10 on Creator, 12 on Pro, and 15 on Teams. See thedeck-upload guide for the current limits and supported formats.
Worked example: a 10-slide sales pitch
Suppose the deck contains a problem, product, proof, pricing, and next step. The narration should give the audience the explanation that normally happens in the room:
- Title slide: state what the pitch will help the buyer decide.
- Problem slide: explain the consequence of the problem instead of reading the bullets word for word.
- Product slide: connect the product to the problem already established.
- Proof slide: call out the one metric or customer result that supports the argument, and verify that it is current.
- Pricing and ask: explain what the buyer should do next and where to get answers.
If a generated line repeats three bullets, rewrite it to say what those bullets collectively mean. If a slide has no useful context beyond its visible text, keep the narration short.
Write narration that adds to the slide
Use the slide for proof
The audience can read a headline, label, or number. The narration should explain why it is present, how it connects to the previous slide, and what decision it supports.
Use one idea per slide
If a slide needs a long explanation, either simplify the slide or split it. A video can hold a slide for the length of its narration, but that does not make a dense slide easier to understand.
Say the context you normally provide live
Add the caveat, example, definition, or transition you would say in the room. That context is usually the part a static deck loses when it is emailed without its presenter.
Make the video usable after export
- Add captions for viewers who watch without sound.
- Use chapters when the deck contains sections that viewers may want to revisit.
- Keep the original deck and the narrated video versioned together.
- Keep the written handout beside the video for search, printing, and quick reference.
- Review prices, dates, metrics, and product screenshots before resending the video.
A narrated deck can serve a sales follow-up, investor update, training lesson, course module, or internal announcement. Change the opening and call to action for the audience rather than sending the same pitch unchanged to every group.
Inputs, outputs, and current limits
- Inputs
- PPT, PPTX, PDF, or a Google Slides/Keynote export
- Deck size
- Up to 100 slides and 650 MB per upload
- Video length
- 6 minutes Free, 10 Creator, 12 Pro, 15 Teams
- Output
- Narrated MP4 plus a written handout with each slide and its script
- Languages
- 80+ from the same deck, without re-recording
- Free plan
- 8 minutes of narration per month; free exports carry a watermark
Common mistakes
- Reading every bullet aloud while the viewer is trying to read it.
- Putting several unrelated ideas on one slide.
- Using a first draft without checking figures, names, and product terminology.
- Choosing a voice that does not fit the audience or the subject.
- Leaving an old price, date, or screenshot in a deck used repeatedly.
- Publishing only the video and losing the searchable written explanation.
Frequently asked questions
What file types can I upload?
ScreenDub accepts PDF, PPT, and PPTX presentation files. Export Google Slides or Keynote to PowerPoint or PDF first.
Can I use a Google Slides deck?
Yes. Export the deck from Google Slides as PowerPoint or PDF, then upload that file.
Does ScreenDub read the content on each slide?
It analyzes the slide content and drafts narration for each slide. Review the draft carefully, especially figures, names, and technical terms, before exporting.
How long can the narrated video be?
The per-video limit is 6 minutes on Free, 10 on Creator, 12 on Pro, and 15 on Teams. A deck can contain up to 100 slides and be up to 650 MB.
Can I use my own voice?
Yes. ScreenDub can use a microphone recording, or you can choose an AI voice and edit the narration as text before exporting.
Can I narrate the same deck in another language?
Yes. Translate and re-narrate the same project into 80+ languages while keeping the original slides and timing.
Do I get anything besides the video?
Yes. The project also produces a written handout pairing each slide with its script. Captions and chapters are available as finishing options.
Start with a deck you already present repeatedly. Create a free ScreenDub account, or see the full PowerPoint-to-video workflow.