How to Do Voiceover on Instagram: 3 Methods

Short answer: You can add a voiceover to an Instagram Reel in three ways: record it in the Reel editor with the Voiceover tool, bring a finished audio file in through the Edits app, or generate the narration on your Mac and import that file. The in-app route is fastest, the Edits route gives you a separate track you can edit, and the generated route gives you a reusable voice and a script you can revise without re-recording.

This guide quotes the Instagram Help Center pages as they read on 2026-10-19, because menu names and feature availability change. It covers Reels first, since Instagram's own voiceover article is written for Reels, then explains the Story path and the one place the documentation is thin. It also covers script length, export formats, and the AI label that realistic synthetic narration now requires.

The three methods at a glance

Method 1: Record a voiceover inside Instagram

This is the path most people want, and it takes under a minute once the clip is ready. Instagram's help page describes it in five steps:

1. Record a clip for your reel, or select a clip from your gallery.
2. Tap Edit Video at the bottom of the screen.
3. Tap Voiceover at the bottom.
4. Tap or hold Record to record audio over the video.
5. Tap Done, then finish sharing your reel.

Three details on that page matter in practice. You can record more than one voiceover clip for the same reel, so a long narration can be built in pieces. A slider lets you set where each voiceover starts, which is how you line a line up with a cut. To remove a take, tap the voiceover track and then tap Discard.

The same page explains attribution. If the reel uses licensed audio or another creator's original audio, only your voiceover is attributed to you and appears next to your username. If the reel uses only your original audio, the audio page for the reel is attributed to you and includes the voiceover. That matters if you plan to let other people reuse the sound.

What about Stories

Instagram's Story help is narrower. The page on sharing a photo or video to your story lists Draw, Text, and Sticker as the editing tools, and the voiceover article is filed under Reels. So for a Story you have two documented options: speak while the camera records, which puts your voice into the clip's own audio, or build the video in Edits and share the exported file. A Story video up to 60 seconds long appears as one clip; longer videos are broken into multiple clips and can be trimmed in the app on iOS.

Method 2: Import a pre-made audio track

Instagram's Reel audio help page lists three sources: your own original audio, original audio from another Reels creator, and the Instagram audio library. It does not describe picking a file out of the Files app inside the Reel editor. The documented file import lives in Edits, Instagram's separate editing app, and that is the route to use for a narration file you already made. If you also post to TikTok, the TikTok voiceover workflow uses the same generate, export, import sequence with a different upload screen.

1. Open the project in Edits and tap Audio.
2. Tap Collections.
3. Tap Import from files to bring in an audio file, or tap Extract from video to pull the audio out of a video on your phone.
4. Trim and place the track. Edits offers Split, Volume, Fade audio, Volume ducking, Audio FX, Beats, Replace, Copy, Duplicate, and Slip for a selected audio layer.
5. Tap Next, then Download to save the video to your camera roll, or Share to Instagram to open it as a reel.

Two limits are worth knowing before you start. An exported video is a snapshot: if you change the project afterward, the file on your camera roll does not update, and you have to export again. And if the imported audio is licensed music rather than your own recording, Instagram may mute it when its agreements with rights holders change, and a downloaded reel that uses licensed audio has that audio removed.

Method 3: Generate an AI voiceover and bring it in

The generated route makes sense when the same voice has to carry a series, when the script contains something you would rather not paste into a phone app, or when you want to fix one mispronounced word by editing text instead of recording the whole take again. The workflow is the same as any imported track, with one extra step at the end for disclosure.

1. Write the script in timed blocks. Put the hook in the first two seconds and give each visual change its own line.
2. Generate the speech in a desktop app and listen to the whole file, not just the first sentence.
3. Export a WAV master and, if you want a smaller file for transfer, an M4A copy.
4. Move the file to the phone with AirDrop or a cloud folder.
5. Import it in Edits through Audio, Collections, Import from files, or place it in a desktop video editor and export the finished video.
6. Share the result as a Reel or Story, then turn on the AI label described below.

Murmur is one option for step two. It is a local voice studio for Apple Silicon and Intel Macs running macOS 15 or later, and it runs the models on the machine rather than sending the script to a cloud service. Kokoro is bundled; the app can also install Chatterbox Turbo, Chatterbox Multilingual, Qwen3-TTS Base, Voice Design, SparkTTS, Fish Audio S2 Pro, Breeze TTS 2, and OmniVoice. It exports WAV and M4A, not MP3, so convert a copy if a tool insists on MP3. The free plan covers Kokoro and scripts up to 1,500 characters, and Pro is $49 one-time with a 3-day trial and a 7-day refund policy. If you are comparing tools first, the Mac AI voice generator guide covers the tradeoffs, and the WAV, M4A, and MP3 export guide explains why the master and the delivery copy should not be the same file.

Voice cloning needs a firm rule: clone only your own voice or a voice you have written permission to use. A generated voice should never be used to imply that a real person said something they did not say. That is a legal and ethical line, not a setting in the app.

Script length for a 15, 30, 60, or 90 second Reel

Use 150 words per minute as a planning rule of thumb, which is 2.5 words per second. It is not a measurement of your voice or your edit, and it leaves no room for pauses, music hits, or time spent letting a visual land. Read the draft aloud with a timer before you generate or record it.

Keep the platform's own limits separate from your script target. Instagram's page on recording a reel says you can record and edit videos up to 20 minutes, and that reels over 3 minutes will not be recommended to new audiences. That is a statement about upload and distribution, not advice to fill the time. Most talking Reels work better well under a minute.

Which method should you use

• Use the in-app recorder when the clip is already on your phone and you only need one pass of narration.
• Use Edits when you want to move the voiceover, cut it, duck music under it, or replace one section without re-recording the rest.
• Use a generated voice when you need the same voice across many posts, a language or accent you cannot perform, or a script you will revise several times.
• Skip all three and record on a separate microphone if the voice is the main product, such as an audiobook sample or a paid ad.

Label a realistic AI voiceover

Instagram's help page on labelling AI content is explicit. Meta requires a label on content that has photorealistic video or realistic-sounding audio that was digitally generated or altered, including with AI. The page's own examples include a reel narrated with a realistic AI-generated voiceover, and it warns that there may be penalties for not labelling content as required.

1. For a Reel: create the reel, then tap the toggle next to Add AI label before sharing.
2. For a Story: share the photo or video to your story, tap Options, tap Label AI, then tap the toggle next to Add AI label.
3. For a profile label: go to your profile, tap Edit profile, then turn on AI-generated profile.

The page also notes that if you choose not to identify AI use, a label may still be applied when Meta detects that the content was AI-generated or AI-edited, and that not all AI content contains the information needed to identify it. Do not strip metadata to dodge a label. Add the label yourself, and consider a short line in the caption as well, because the label is easy to miss on a fast scroll. The AI voice disclosure guide covers the same habit on another platform.

Instagram also has its own text-to-speech for Reels, which reads text you type with a small set of voices. Its help page says that feature is currently English only and limited to select countries where captions are available. That tool is fine for a quick on-screen caption read. It is not a substitute for a recorded or generated narration file when you need a specific voice, a longer script, or a language outside that list.

Problems you will hit, and the fix

• The narration is buried under the clip's own sound: lower or duck the competing audio before you share. Edits offers Volume and Volume ducking for its audio layers.
• The voiceover sounds roomy: the phone microphone picks up the room, so record closer or use an external microphone.
• An imported file will not appear: confirm the file is on the phone and that you used Import from files, not Extract from video.
• Your Edits changes are missing from the posted video: export again, because the earlier file is a snapshot.
• A licensed track went silent: Instagram can mute audio when rights agreements change, so keep a copy of your own narration file.
• A tool rejects the export: Murmur writes WAV and M4A. Convert a copy to MP3 rather than renaming the extension.

Frequently asked questions