How to Add AI Narration to PowerPoint and Keynote on Mac (Step-by-Step)

A slide deck can explain a product, teach a lesson, or introduce a project without a live presenter. The challenge is giving every slide a clear voice without recording yourself over and over.

On a Mac, you can generate AI narration from a script, save it as an audio file, and add the audio to PowerPoint or Keynote. The workflow works particularly well for employee training, asynchronous presentations, narrated lessons, and product walkthroughs.

This guide uses Spokio for Mac for local English text-to-speech. You create the recording in Spokio, then insert an MP3 or another compatible audio file into your presentation. Spokio does not have a direct PowerPoint or Keynote plugin.

There is one important difference between the apps: PowerPoint for Mac can play imported audio in a slideshow, but Microsoft’s video export can omit that imported audio. Keynote has a separate Movie export workflow that can include audio. We’ll cover both, including how to make a video when you need one.

Quick answer: Add AI narration to a presentation on Mac

For PowerPoint:

Write a short narration script for each slide.
Generate the spoken sections with Spokio → Create Audio → Quick Audio.
Export each approved section as MP3.
In PowerPoint for Mac, select a slide and choose Insert → Audio → Audio from File.
Choose the corresponding MP3, then set Playback → Start → Automatically.
Repeat for the other slides, adjust transitions and timings, and test the slideshow.
If you need an MP4 with sound, see the separate workaround below—don’t assume PowerPoint’s video export will include the imported MP3s.

For Keynote:

Generate and export a narration file for each slide in Spokio.
Open Keynote and drag the audio file from Finder onto its slide.
Use the slide’s audio playback settings to start the file when the slide appears.
Set slide and build timings so narration finishes before the next transition.
Test the presentation. For a video, choose Movie export and verify the resulting MOV plays with sound.

Both methods start with an ordinary audio file, not a special AI presentation format.

Should you use PowerPoint, Keynote, or a video editor?

Goal	Good approach	Caveat
Editable slides that play AI narration	PowerPoint + inserted audio	Test playback in the actual slideshow
Editable Apple presentation with AI narration	Keynote + inserted audio	Set each slide’s start and transition timing
Narrated MOV directly from a Keynote deck	Keynote Movie export	Preview the actual exported movie
Narrated MP4 from PowerPoint for Mac	Export slides to video, then combine with audio in an editor	Imported audio may be absent from native video export
Fully polished tutorial with tight voice-to-visual edits	Export visual scenes and finish in a video editor	More editing, but precise control over pacing

The key question is whether your audience will open the presentation or watch a finished video. Those are not always equivalent workflows.

Step 1: Write your narration slide by slide

Good presentation narration is not a recording of every bullet. Your audience can read the slide already; the voice should connect the ideas and supply useful context.

For a short product presentation, you could plan:

Slide	Visual purpose	Narration purpose
1. Introduction	Show the topic and hook	Explain the problem in one or two sentences
2. How it works	Show the workflow diagram	Walk through the important steps
3. Benefits	Show evidence or examples	Interpret what the figures mean
4. Next step	Show a clear action	Tell the viewer what to do next

Keep each slide’s narration as its own script. One option is to maintain a folder like this:

presentation/
  scripts/
    01-introduction.txt
    02-workflow.txt
    03-benefits.txt
    04-next-step.txt
  narration/
    01-introduction.mp3
    02-workflow.mp3
    03-benefits.mp3
    04-next-step.mp3

This is just an organizational example. Spokio does not require those filenames.

How long should a slide stay on screen?

As a starting estimate, natural English narration often runs around 130–160 spoken words per minute. A script of about 50 words might take roughly 19–23 seconds, excluding pauses.

But always measure the actual exported audio. A slower voice, technical terms, or more punctuation may change the duration.

If a slide’s audio runs 23 seconds, give that slide enough time to finish, plus a small visual pause if it helps. When you use builds or animations, allow time for those visual events too.

Step 2: Generate slide narration locally in Spokio

The simplest approach for a deck with a few slides is Quick Audio.

Open Spokio for Mac.
Choose Create Audio → Quick Audio.
Paste the script for the first slide.
Choose an English voice and adjust its Speed.
Generate the recording.
Listen before exporting. Fix names, numbers, and awkward wording.
Choose Export → Export as MP3, and save the result.
Repeat for each slide.

Spokio performs normal speech generation on your Mac rather than sending the script to a cloud TTS API. You can use built-in voices or a voice sample you have permission to use. The source presentation might still be stored in OneDrive, iCloud Drive, or another sync service; local speech generation does not change that.

Keep the voice consistent across slides

Choose a voice and speaking speed early. Changing voices in the middle of a deck can make a lesson feel like several unrelated recordings.

Generate one representative slide before committing to the entire deck. Test the pronunciation of acronyms, company names, percentages, and other terms you expect to repeat.

If you run into errors, see How to Fix AI Voice Pronunciation: Names, Acronyms, and Numbers.

Should I use MP3, WAV, or M4A?

Format	Useful when	Considerations
MP3	You want small, widely compatible slide narration files	Good default for PowerPoint and most presentation workflows
M4A	You want compact Apple-friendly audio	Useful fallback for Keynote if an imported file causes trouble
WAV	You expect further audio editing	Larger, uncompressed editing master

Spokio can export MP3, M4A, WAV, and AIFF for Quick Audio. PowerPoint’s media format guide lists several supported audio formats; actual compatibility depends on encoding and app version.

For Keynote, Apple recommends audio that QuickTime can play and suggests conversion to M4A if a file can’t be imported. See Apple’s media guide.

Free and Pro limits

For a short narration clip, Spokio Free can be enough. Quick Audio supports up to 1,000 characters per synthesis on Free and 20,000 on Pro.

For many slides, or when you prefer to process multiple individual script files together, Spokio Pro’s Bulk Project supports project generation and separate-file export. A Bulk Project treats each imported source file as an item; it does not directly read PowerPoint slides or automatically convert slide notes into separate audio tracks.

See How to Batch Export AI Voiceovers on Mac for a production-oriented file workflow.

Method 1: Add AI voiceover to PowerPoint on Mac

PowerPoint for Mac can insert a saved audio file onto a slide. Once you’ve generated MP3 recordings in Spokio, you no longer need the text-to-speech app open while presenting.

Step 1: Insert the audio file

Open your .pptx in Microsoft PowerPoint for Mac.
Select the slide that should play the narration.
Choose Insert → Audio → Audio from File.
Select the corresponding exported MP3.
Confirm the audio icon appears on that slide.

Menu wording may vary slightly between PowerPoint versions. Microsoft’s instructions for adding audio distinguish Mac’s Audio from File option from the Windows wording Audio on My PC.

Do not simply paste text into the slide and assume PowerPoint will synthesize it. In this workflow, the AI voice is the audio file you explicitly insert.

Step 2: Make the audio start automatically

If viewers must click the audio icon to hear the voice, an unattended slide deck will feel broken.

Select the inserted audio icon.
Open PowerPoint’s Playback tab.
Set Start → Automatically.
Optionally enable Hide During Show, so the speaker icon does not clutter the slide.
Test the slide using Slide Show.

PowerPoint also offers Play Across Slides, but that is better for a continuous soundtrack than individual slide narration. For the clearest slide-by-slide workflow, keep one narration file attached to the appropriate slide.

Step 3: Match slide duration to narration

If a slide advances before its MP3 finishes, the audience may miss the end of a sentence. Review the slide’s Transitions and timing settings, and test the actual slide show.

For self-running presentations:

Set the transition to advance after the recording has had time to finish.
Account for animations or builds that consume part of the slide’s total duration.
Avoid automatic transitions that cut off long audio clips.
Test a three-slide sequence rather than only one slide in isolation.

For a live presentation, manual advancement is often easier; the presenter can advance only after narration ends.

Step 4: Repeat and test the whole deck

Import the corresponding MP3 for each slide. Play the deck from the beginning and listen through every transition.

Look for:

Two recordings overlapping
The next slide appearing while the previous narration continues
A silent slide with no audio
A clip starting only after an accidental click
Different volumes or speaking speeds between slides

Save a copy of your edited presentation before experimenting with new timing or media settings.

Important: PowerPoint for Mac video export can omit imported narration

This is the most important limitation in this article.

PowerPoint’s slideshow playback and video export do not handle all audio the same way. Microsoft’s PowerPoint for Mac video export documentation states that video export can include narration recorded using Slide Show, but other embedded media files don’t play in the exported video.

That means:

An MP3 inserted onto a slide may play correctly when you run the deck in PowerPoint.
The same file may be missing from an MP4 exported by PowerPoint for Mac.
Selecting Use Recorded Timings and Narrations is not a guarantee that separately inserted AI narration will be included.

Do not promise yourself—or a client—that a PowerPoint export contains the AI narration until you’ve watched the resulting MP4 with its sound on.

Reliable alternative for a narrated MP4

If you need a shareable video with an AI voice track, use PowerPoint for the visuals and a video editor for the final audio.

In PowerPoint for Mac, set the slide timings for the intended visual sequence.
Choose File → Export → MP4 to produce the video.
Open that exported video in a video editor such as iMovie or Final Cut Pro.
Import the Spokio narration files as audio tracks.
Align each audio clip to the corresponding slide, scene, or visual cue.
Trim the video scenes or adjust their lengths to fit the spoken clips.
Export the finished MP4 from the video editor.
Play the exported file all the way through to verify the audio.

The native PowerPoint video export also has its own limitations around transitions and animations, so inspect the visual result before mixing audio. For a detailed audio-to-video editing workflow, see How to Create AI Voiceovers for Final Cut Pro on Mac.

If you’d rather narrate directly inside PowerPoint using a microphone, its Slide Show recording feature is a different workflow and may be included in video export as Microsoft’s documentation describes. Playing an AI MP3 near a microphone to simulate live narration is not the same as a clean imported-audio workflow and can introduce recording-quality problems.

Method 2: Add AI narration to Keynote on Mac

Keynote is a good choice if you build presentations in Apple’s ecosystem and want audio files attached to individual slides.

Step 1: Add the exported audio to a slide

Open the presentation in Keynote.
Select the first slide you want to narrate.
Find the Spokio MP3 in Finder.
Drag the audio file onto the slide. You can also use Keynote’s Insert Media menu to choose audio.
Select the inserted audio object to review its playback settings.

Apple documents this under Add video and audio in Keynote on Mac.

If the file won’t import, first test it in QuickTime Player. Apple’s guide recommends using a compatible format such as M4A where needed; you can re-export from Spokio in that format.

Step 2: Configure automatic audio playback

In Keynote, slide media can be set to play after a click or to start when the slide appears. You usually want the latter for a self-playing presentation.

Select the audio object on its slide.
Open the Format → Audio sidebar.
Find the audio start/playback option.
Configure the clip to start when the slide appears rather than waiting for the presenter to click it.
Present the slide and confirm that the audio starts as intended.

The exact control labels can vary with Keynote versions. Apple’s audio editing guide covers start behavior, trimming, volume, and audio playing across slide transitions.

Step 3: Choose per-slide narration or a continuous soundtrack

Keynote supports different audio layouts:

Setup	Better for	How to think about it
One clip per slide	Training, lessons, modular product demos	Easy to revise only the affected slide
Audio across consecutive slides	A sentence spanning two or more slides	Keep the same audio playing across the intended slides
Presentation-wide soundtrack	Background music or one uninterrupted narration	Harder to revise or align individual slide narration
Record Slideshow	Recording yourself while changing slides	Not a direct replacement for inserting Spokio audio files

For a course that will be revised later, one narration clip per slide is a practical default.

Apple also documents playing audio across slides when an uninterrupted track is necessary. Do not enable looping for ordinary narration unless repetition is genuinely intended.

Step 4: Set slide and build timings

Audio doesn’t automatically guarantee that the presentation changes slides at the right word. Configure the deck’s transition behavior and test any builds.

For each slide:

Listen to the MP3 and note its actual length.
Set the slide’s advance timing to give that audio enough time.
Check whether builds appear at sensible moments.
Preview the complete sequence with narration.

If a slide contains 25 seconds of spoken explanation but advances after five seconds, increasing narration speed usually isn’t the right fix. Give the slide enough time or shorten the script and generate a replacement.

Step 5: Export Keynote with narration as a movie

If your audience needs a video rather than a Keynote file, Apple offers a Movie export.

Save your Keynote presentation.
Open the export options, choosing Movie as the output format.
Select a resolution and self-playing timing settings appropriate to your project.
Export the movie.
Open the resulting MOV in QuickTime Player and check that all narrated slides have sound.

Apple’s Keynote export instructions explain the available export formats. Apple’s send/export guide describes Movie as a .mov output that can include audio. If you used Keynote’s own recorded slideshow feature, choose the Slideshow Recording playback option as described by Apple.

If you specifically need MP4, note that Keynote’s Movie export is typically MOV. Use a video editor or converter to create the desired final format, and verify sound after conversion.

PowerPoint vs. Keynote: AI narration workflow compared

Capability	PowerPoint for Mac	Keynote for Mac
Insert narration files generated in Spokio	Yes	Yes
Play narration automatically on a slide	Yes	Yes
Keep an editable presentation with audio	Yes	Yes
Use one narration clip per slide	Yes	Yes
Native movie export reliably includes inserted clips	Don’t assume so; Microsoft documents exclusions for embedded media	Apple’s Movie export supports including audio; test your exported file
Typical native video output	MP4 or MOV, depending on version	MOV
Best route to a polished narrated MP4	Export visuals, combine narration in a video editor	Export Movie and convert/edit when needed

Neither PowerPoint nor Keynote reads a Spokio project file directly. The shared bridge is an exported audio file.

Editing one slide without rerecording the entire presentation

One major advantage of generating speech from written text is how small a revision can be.

Imagine slide 6 contains a product name that changed.

Open the slide 6 script in your text editor or Spokio.
Replace the old name and listen to a fresh generation.
Export a revised file, for example 06-product-update-v2.mp3.
In PowerPoint, delete the old audio object and insert the new file on slide 6.
In Keynote, select the existing audio object and use its Replace control, or insert the new file and remove the old object.
Recheck that slide’s duration and playback behavior.
Preview slide 5 through slide 7 to catch boundary problems.

You do not have to regenerate unrelated slides. For bigger workloads, keeping narration in independent clips also helps when different lessons or modules are updated throughout the year.

Privacy, source rights, and voice cloning

Spokio generates the voiceover locally on your Mac. That means the normal speech synthesis step doesn’t require uploading the script to a remote speech service. This can be helpful for internal training decks or unpublished product presentations.

However, if a deck is stored in OneDrive, iCloud Drive, or a collaboration tool, its contents and any shared audio files may be synced there. Local text-to-speech processing is not a statement that your entire presentation workflow is offline.

If you clone a voice, use only a sample you own or have explicit permission to use. Publishing an AI narration can also raise questions about rights to the script, logos, diagrams, and other content inside the presentation. Check those rights before distributing a narrated course or client deck.

Troubleshooting PowerPoint and Keynote voiceovers

The narration plays only when I click it

In PowerPoint, select the audio icon and check Playback → Start → Automatically. In Keynote, inspect the audio playback options and set the clip to begin with the slide. Test in slideshow mode, not just while editing.

My audio stops before the slide is finished

Confirm that the media file itself plays through to the end. Then review trimming settings, playback controls, and slide timing. If a slide changes too soon, lengthen its duration or rework the narration.

Two slides speak at the same time

Check whether a clip is set to play across slides while the next slide has its own audio. Disable unintended across-slide playback, or redesign the sequence so only the intended narration continues.

My PowerPoint MP4 is silent, but the slideshow has audio

This is a documented PowerPoint for Mac export limitation for embedded media. Microsoft’s video export supports narrated audio recorded through Slide Show but does not promise that inserted MP3s will play in the exported video. Use the export visuals → combine narration in a video editor method above.

Keynote won’t accept my audio file

Test the file locally, then export another copy in M4A from Spokio. Apple says Keynote media must be compatible with QuickTime and suggests converting files that are not supported.

Keynote’s MOV has sound but the timing is wrong

Review self-playing transition intervals and builds. The export settings and slideshow’s playback behavior can differ from how you advance slides manually. Re-export after correcting the timing and listen again.

A name or acronym sounds wrong

Edit the narration text in Spokio and regenerate the corresponding slide’s clip. For more examples, see How to Fix AI Voice Pronunciation.

I changed my presentation notes, but the AI audio didn’t change

The exported audio is a snapshot of the script at generation time. Editing PowerPoint or Keynote notes does not automatically update files you already generated in Spokio. Regenerate and replace the affected recording.

Frequently asked questions

Can PowerPoint for Mac play an AI voiceover?

Yes. Insert an MP3 onto a slide and set its playback to Automatically. Test the slideshow to verify that narration and slide timing work together.

Can I save a PowerPoint with AI narration as MP4?

You can export PowerPoint slides as MP4, but the imported audio may not be included in PowerPoint for Mac’s video export. Microsoft distinguishes Slide Show-recorded narration from other embedded media. For a reliable final MP4 with externally generated audio, combine the visuals and narration in a video editor.

Can Keynote use AI-generated audio?

Yes. Add a compatible audio file to a slide, set its playback behavior, and adjust the slide timing. Keynote can export a MOV including audio; verify the finished file.

Does PowerPoint have built-in text-to-speech narration?

PowerPoint offers features for recording spoken narration and system accessibility features for reading text, but recording a slideshow is not the same as creating a reusable AI voiceover file from a script. This guide focuses on generating a separate file with Spokio.

Is there a Spokio plugin for PowerPoint or Keynote?

No. The workflow is generate → export audio → insert into your presentation. Changes are not synchronized automatically between the apps.

Can I use Spokio Free for slide narration?

Yes, if each Quick Audio generation fits the 1,000-character Free input limit. Pro expands individual Quick Audio input to 20,000 characters and offers project-based generation and export for larger sets of files.

Can I use a cloned voice to narrate my entire deck?

Spokio supports local voice cloning from permitted samples. Keep the chosen voice and speed consistent across all slides. Voice cloning does not override the consent or usage rights of the original speaker.

What should I deliver to a client: the slide deck or a video?

If the client needs to edit and present individual slides, an editable PowerPoint or Keynote file makes sense. If they need a consistent viewing experience with sound, deliver a tested narrated video, and keep the editable source file separately.

Make each slide sound as polished as it looks

A repeatable approach is write one short script per slide → generate AI speech locally → export audio → insert it → test timings → deliver the presentation or video.

For PowerPoint on Mac, remember the special MP4 export limitation. For Keynote, use the Movie export and check the resulting MOV. In either app, a few minutes of audio review can prevent silent exports, clipped endings, and awkward pronunciation.

Try Spokio for Mac to create reusable English narration locally for your next lesson, presentation, or product demo.