A finished slide deck often has to do its job without you in the room: a prospect who missed the call, a new hire working through onboarding slides on their own schedule, or a sales deck your reps send out without booking a live walkthrough for every lead. Static slides don’t cover that gap the way a voice over does.
PowerPoint has two built-in ways to record your own narration, slide by slide or in one continuous pass, and that’s still the most direct route to a file that plays your voice back for whoever opens it next. If you’d rather not record yourself but still need that playable file, a couple of AI add-ons can generate a synthetic voice-over and embed it into the deck for you (covered below). And if what you actually want is a fast way to hear a deck talked through while you’re still building it or getting ready to present it live yourself, Zoom Slides, the presentation product inside the Zoom AI Productivity Suite, can generate an AI voice for each slide right inside the editor.
Recording Your Own Voice Over Directly in PowerPoint
Ready to create a deck? Open Zoom Slides.
If you’re comfortable narrating in your own voice, PowerPoint has two built-in ways to capture it, and neither requires a plugin. Microsoft’s own instructions cover both in detail, but here’s the short version. Both are desktop-app features (Windows and Mac); PowerPoint for the web doesn’t include the recording tools.
Method 1: Record the entire slide show in one pass.
- Open your deck and go to the Slide Show tab.
- Click Record Slide Show, then choose whether to record from the current slide or the beginning.
- Confirm your microphone (and camera, if you want your video to appear) is selected, then click Start Recording.
- Talk through each slide as you would in a live presentation. PowerPoint captures your narration along with slide timings, ink annotations, and laser pointer movements, so the deck can advance itself when someone plays it back later.
- Click the stop icon when you’re done, then save the file. Your narration is now embedded slide by slide.
Method 2: Record narration one slide at a time.
- Select the slide you want to narrate.
- Go to Insert > Audio > Record Audio.
- Name the clip, click the record button, and speak your narration for that slide.
- Click stop, then use the play button to preview the take.
- Click OK to insert the audio onto the slide, or Record again if you want another take before accepting it.
- Repeat for each slide that needs narration.
Both methods put a real, playable audio track behind your slides, but they’re not identical. Method 1 also records slide timings, so the deck can advance on its own when someone opens it later. Method 2 just places an audio icon on each slide, and whether that clip starts automatically depends on the playback settings you choose for it afterward.
Either way, you’re doing the recording yourself: a decent microphone, a script or enough familiarity with the content to narrate cleanly, and the willingness to re-record a slide if an edit changes what the narration needs to say. For a one-off internal update, that’s often fine. For a deck you’d rather not narrate yourself but still need to send out with sound attached, the AI voice-over add-ons below fill that gap. If what you’re after instead is a way to hear the deck talked through as you build it, that’s a different job, covered further down in Zoom Slides.
If You’d Rather Not Record Anything Yourself
A handful of tools exist specifically to generate a synthetic voice-over for a finished PowerPoint file, so you never have to touch a microphone. Murf.ai, for example, offers PowerPoint and Google Slides add-ons that let you write or paste a script and sync AI voices to your slides without leaving the editor, or you can work in Murf’s separate studio and export the audio for later.
SlideSpeak takes a different approach: you upload a finished .pptx file, and its AI uses the on-slide text and speaker notes to create a narrated version. Processing time and output behavior can vary, so review the returned file before sharing it.
Both are useful, and both share the same basic pattern: your deck and your voice-over tool live in two separate places, so you export, upload, or install an add-on to get the narration attached after the fact, but the payoff is a file that plays that narration for whoever opens it next, which is the point if you’re sending the deck out. Zoom Slides takes a different angle on the same “I don’t want to record this myself” problem. Its voice-over panel sits inside the same editor where the deck itself was built, so you can hear a slide read back to you the moment you finish writing it: a fit for talking yourself through the content or rehearsing before you present live, rather than for producing a file that plays sound for someone else. If getting a narrated file out the door to a recipient is the goal, Murf and SlideSpeak above (or PowerPoint’s own recording tools) are the more direct route.
Generate an AI Voice-Over Inside Zoom Slides: Built for Talking Through the Deck Yourself
Zoom Slides is the presentation product inside the Zoom AI Productivity Suite. You can start it from a written prompt, from an existing document, or from a Zoom meeting’s discussion, and it builds a full deck, layout, copy, and speaker notes included, that you can keep editing afterward.
For this walkthrough, the deck already exists: a 10-slide sales pitch built from a single prompt describing a mid-market retail client evaluating a SaaS inventory platform, complete with an agenda, the client’s stockout and manual-tracking pain points, a product overview, and a three-tier pricing page. The voice-over step below picks up from there, on a finished deck sitting in the Slides editor.
If you’re starting from a deck you already built somewhere else, you don’t need to recreate it from a prompt to reach this same step. Zoom Slides imports existing .pptx files, and once that deck is open in the editor, the narration bar and voice-over panel below are available on every slide the same way, whether the slide came from a prompt or from a file you brought in.
Once you’re looking at a slide you want narrated, adding the voice-over takes a few clicks:
- Open the slide in the editor. A narration bar sits underneath the slide canvas, and it already pulls in the on-slide text (titles, stats, bullet points) in the order they appear, so you’re not starting from a blank script.
- Next to that bar is a round voice icon. Click it to open the voice-over panel.
- The panel has two tabs, Voiceover and Music, plus a language selector currently set to English.
- Under Voices, pick who reads the slide. The default is Carol, and the list scrolls through several other options: Jessa, Samara X, Hope, Siren, Frederick Surrey, and more, each labeled with a gender, age range, and accent. The roster even includes a preset labeled Burt Reynolds™.
- With a voice selected, click Generate voice. That sends the slide’s text to the voice you picked and starts generating the narration for that page.
- While the clip is generating, the round icon on the narration bar turns into a loading spinner next to a Stop control. There’s no progress bar or time estimate, so the spinner is the only sign it’s working.
- Once the clip is ready, it starts playing back automatically, and the icon switches to a pause symbol: that’s the “done” signal. Click the icon to pause, and it turns into a play triangle you can click again to hear the same clip from the start. The icon shape itself (spinner, then pause, then play) is the signal to watch for whether a slide already has narration; the tooltip label underneath stays fixed on “Generate voice” throughout, so it won’t update to tell you that.
A few things worth knowing about that playback icon before you rely on it: moving to another slide and back within the same working session doesn’t erase what you generated; the icon comes back showing a play button, and clicking it replays the existing clip instantly, with no regenerating. But if you close the file and come back to it after a full page reload, clicking that same icon triggers the generating spinner again rather than instantly resuming a saved clip, so a voice-over you generated today isn’t guaranteed to be instantly ready to play the next time you open the file: you may need to regenerate it. There’s also no badge or waveform on the slide thumbnails in the left-hand panel to show which slides already have narration and which don’t; the only way to check is to open each slide’s narration bar directly.
Zoom Slides uses preset AI voices to generate quick, consistent narration for draft playback and rehearsal.
It’s a narrator-selection feature, not a voice-cloning one: you’re picking a preset AI voice to read the deck, rather than making the narration sound like your own.
That panel opens from the narration bar on whichever slide you’re viewing, so this same choose-a-voice-and-generate flow is what you’ll use slide by slide as you move through the deck. Pick the same voice each time you open the panel if you want one consistent presenter throughout.
Where the Zoom Slides AI Voice-Over Actually Helps Right Now
Not needing a microphone, a quiet room, or the confidence to narrate your own pitch deck out loud is the obvious win, and it shows up fastest when you’re the one who benefits: hearing the deck read back to you while you’re still writing it, running through a pitch out loud before a live demo, or catching a line that sounds fine on the page but awkward out loud, all without touching a mic or re-recording anything by hand.
Zoom Slides describes the narration as designed to autoplay, with slides advancing automatically and narration synced to each one, a capability aimed at decks someone opens without you in the room, like a prospect reviewing pricing after a call or a new hire working through training slides on their own. That autoplay behavior isn’t live in Slideshow view yet: starting Zoom Slides’ own Slideshow on a slide with narration already generated, both as the deck’s owner and after opening the same deck through a shared “anyone with the link” copy, plays no narration out loud and shows no audio control on screen. Until that lands, the safer way to use this feature today is as something you listen to yourself inside the editor, not as a deck you hand off already narrating itself.
It also holds up better through revisions. Record your own narration, then edit the slide text, and the audio no longer matches; you’re back in front of the microphone for that slide. With an AI voice-over pulling from the on-slide text, an edit that changes what a slide needs to say means regenerating that slide’s narration instead of re-recording it yourself.
Getting the deck itself into a standard file format isn’t a separate hurdle, either. Zoom Slides supports importing and exporting .pptx files. Layouts, fonts, animations, and other complex elements may vary during conversion, so review the deck in PowerPoint before use. The narration is the part that doesn’t make that trip, as the next paragraph covers.
In this walkthrough, the AI narration track did not carry over into the exported .pptx: no audio media was present, whether or not narration had been generated for the slides beforehand. Visual formatting and other elements should be reviewed separately after export. Sharing the deck from inside Zoom Slides didn’t clear that hurdle either: opening a shared link as someone without edit access showed the on-slide script text but not the generate or play control available to the deck’s owner.
In other words, hearing the AI voice-over right now means listening to it inside the Zoom Slides editor itself, with edit-level access to the file. Neither the exported .pptx nor a shared view-only link carries the narration through to a recipient. If a file that narrates itself for someone else is what you need, that’s what PowerPoint’s built-in recording tools (or an AI add-on that embeds the audio, like the ones covered earlier) are for, at least until that gap closes.
FAQ
Do I have to record my own voice, or can I use an AI voice instead? You don’t have to record anything: Zoom Slides generates narration using a preset AI voice you pick from a list, rather than recording or cloning your own voice. If you specifically need the narration to sound like you, PowerPoint’s built-in Record Slide Show or Record Audio tools are the more direct route.
Can I still get a PowerPoint file after adding an AI voice-over? The deck itself, yes: Zoom Slides supports .pptx import and export. Layouts, fonts, animations, and other complex elements may vary during conversion, so review the exported deck in PowerPoint. In this walkthrough, the exported .pptx contained no AI narration audio. Sharing the deck via a link also did not make the narration available to a viewer without edit access. If you need a file that narrates itself for someone else, use PowerPoint’s recording tools or an AI voice-over add-on that embeds audio, then review the resulting file before distribution.
Does the voice-over apply to the whole deck at once, or one slide at a time? The voice-over panel works slide by slide: you open the panel on each slide, choose a voice, and click Generate voice for that page’s narration.
How do I know a slide’s narration actually generated, and can I hear it before moving on? Watch the icon shape on the narration bar, not its tooltip label, which stays fixed on “Generate voice”: the icon turns into a loading spinner while the clip is generating, then switches to a pause icon as the finished clip starts playing automatically. Click it to pause, and it becomes a play button you can click again to replay the same clip. That replay works instantly as long as you stay in the same working session; after a full page reload, clicking it starts the generating process over rather than resuming a saved clip.
Can I add an AI voice-over to a PowerPoint file I already built elsewhere? Zoom Slides can import existing .pptx files, and once the deck is open in the Slides editor, the narration bar and voice-over panel are available on each slide. Review the imported layout and formatting before generating narration.