Comu Note Pro and Comu Action Pro work as an AI recorder that recognizes speakers by voice. You record a 20 to 30 second Voice Profile in the Comu app, and your voice is labeled Me in new recordings. Speaker separation, Name Speakers and Voice Profile come with the free Starter plan (checked Nov 9, 2026).
Both devices use the same Comu app, so the steps below are identical on Comu Note Pro and Comu Action Pro. To set up your voice, jump to the Voice Profile steps. To fix a label that reads "Speaker 1", go to the naming workflow.
What is the difference between speaker identification and speaker separation?
Speaker separation splits a recording into turns and marks which turns belong to the same voice. Speaker identification goes one step further and attaches a real name to each voice. A transcript with separation alone reads "Speaker 1", "Speaker 2". A transcript with identification reads "Me" and the names you entered.
Voice signature speaker identification: what is a Voice Profile?
A voice signature is the stored description of how a voice sounds. When a recorder keeps a voice signature for you, it can compare each new recording against it and label your turns without asking. In Comu, that stored sample is called a Voice Profile.
| Term | What it does | Where it lives in Comu |
|---|---|---|
| Speaker separation | Splits the transcript into voices and shows who spoke when | Automatic on every transcript, with playback-synced highlighting |
| Voice signature (Voice Profile) | Stores a sample of your voice so the app can recognize it in later recordings | Me tab in the app, optional, takes 20 to 30 seconds |
| Speaker identification | Puts a name on each voice in meeting notes | Voice Profile matching labels you as Me; the Name Speakers button names everyone else |
Speaker separation is software, so it works the same on both devices and on every plan. The hardware affects the audio that goes in, not the labeling steps that follow.
How do you set up a Voice Profile in the Comu app?
You set up a Voice Profile by opening Set up Voice Profile in the Comu app, agreeing to the privacy prompt, allowing microphone access and reading a short script for 20 to 30 seconds. The app saves the profile and shows a completion screen that reads Enabled.
The sample is recorded with your phone's microphone.
Steps:
- Open the Comu app and start from any entry point for Voice Profile: the Me tab, the Get Started checklist or the banner shown to new users on the Files page.
- Read the Set up Voice Profile screen, which explains what the profile does, and tap Set up.
- Review the Privacy Policy prompt and tap I Agree. Cancel closes the prompt.
- Allow microphone access. If the app already has permission, this step is skipped and you go straight to the recording screen. If it has not asked yet, the Allow microphone access screen appears. Tap Enable microphone and confirm in the system prompt.
- Read the on-screen script and speak for 20 to 30 seconds. The recording screen shows a timer and the record button.
- Watch for the orange banner. It appears when the app hears no voice, a very low volume or very fast speech, and it does not stop the recording. Adjust and keep going.
- Stop the recording. The app saves your Voice Profile. A save that finishes quickly shows a check mark. A longer save shows a Saving… overlay until it succeeds or fails.
- Read the completion screen. It shows Enabled and says the profile applies to new recordings, and a How it helps section sits below it.
What if microphone access is denied or the save fails?
If microphone access is denied, a fallback screen offers Enable again or Open Settings, and if the save fails, a red banner offers Retry.
| Situation | What you see | What to do |
|---|---|---|
| Microphone access denied once | A fallback screen with Enable again and Not now | Tap Enable again to request access once more, or Not now to leave |
| Microphone access denied permanently | A screen with Open Settings and Not now | Tap Open Settings, switch on microphone access for Comu, then return to the app |
| Sample too quiet, silent or too fast | An orange warning banner during recording | Speak up or slow down; the recording continues |
| Save fails | A red banner with Retry | Tap Retry. The app tries the save again and does not ask you to record again |
| Recording fails | A Recording failed message | Tap Re-record to capture a new sample |
Orange banners are quality hints. Red banners need an action from you.
Can you skip, re-record or delete a Voice Profile?
Yes to all three. Voice Profile is optional, so the rest of the app, including transcription, speaker separation and Name Speakers, works without one. When you want to change it, open the Voice Profile screen, tap the three-dot menu and choose Re-record to capture a new sample or Delete to remove the profile.
Delete asks for confirmation. After you confirm, the Voice Profile entry returns to Not set, and you can run the full setup flow again whenever you like.
How do you name speakers and fix "Speaker 1" later?
You name speakers by opening a recording in Comu and tapping Name Speakers. The button is always visible on the recording detail page, whether or not every speaker already has a name. Voice Profile matching labels you as Me, and Name Speakers is where you rename every other voice or correct a label.
The workflow below works on any finished transcript, including recordings made before you created a Voice Profile.
- Open the recording after the transcript has been generated.
- Tap Name Speakers. A dialog lists each voice found in the recording. Me appears in the list when the app matched your Voice Profile.
- Type a name for each voice, or choose Me for your own voice.
- Confirm. The names update in the transcript and in the labels inside the note text, such as the summary and insights.
The names in the note body are labels, not plain text. That is why one rename in the Name Speakers dialog changes every place the name appears. To edit a name, return to the Name Speakers dialog instead of typing over the label in the note. In the editing view for a summary or insights, a label can be deleted and a name typed in by hand.
How do you fix one wrong label inside the transcript?
Open the transcript, tap the speaker label on the segment, and type the right name. A choice then appears with two options:
| Option | Effect | Default |
|---|---|---|
| Apply to this segment only | The new name applies to the tapped segment and leaves the other segments from that voice alone | Not selected |
| Apply to all segments from this speaker | The new name applies to every segment from that voice, and the Name Speakers dialog and the note text update to match | Selected |
Use the segment-only option when one turn was assigned to the wrong person. Use the all-segments option when a voice carries the wrong name from the start, such as a "Speaker 2" whose real name you now know.
Can you set the number of speakers before the transcript is made?
Yes. The information window that appears before a transcript is generated has a Number of Speakers field. It is a required single choice with Auto as the default, and the other choices are 1, 2 and 3+. The same window has a Recording Language field with Auto as the default and 113 languages to choose from. If you know a recording has two voices, choose 2 instead of Auto.
How do you report a bad speaker label?
The thumbs-down feedback on the outline and the transcript offers seven reasons, and one of them is "unreliable speaker identification". Pick it when a label is wrong.
Do speaker names carry into exported transcripts?
The transcript export window has an Include Speaker Names option, and it starts unchecked. Switch it on when you want names in the exported file. Include Timestamps is the companion option, and it starts checked.
What happens to new recordings and to old ones?
A Voice Profile applies to new recordings. After setup, the completion screen reads Enabled and states that the profile applies to new recordings, so recordings you make from then on are matched against your voice during processing.
For an older recording, open it and use Name Speakers, which works the same way as on a new one.
| Action | What happens |
|---|---|
| Set up a Voice Profile | New recordings are processed with your voice sample, and your turns are labeled Me |
| Name Speakers on any recording, old or new | The names update in that recording's transcript and note text |
| Re-record the Voice Profile | You record a new 20 to 30 second sample from the three-dot menu |
| Delete the Voice Profile | The entry returns to Not set; you can set it up again from the beginning |
| Skip the Voice Profile | Speaker separation and Name Speakers keep working on every recording |
Does Comu support speaker identification across meetings? What Starter and Premium or Elite include
Comu identifies you across meetings with your Voice Profile on the free Starter plan. Matching other people by their voices across meetings is a Premium or Elite feature. Subscriptions are bought inside the Comu app, through the App Store or Google Play.
| Capability | Plan needed |
|---|---|
| Speaker separation with playback-synced highlighting | Starter (free with the device) |
| Voice Profile: set up, re-record, delete | Starter (free with the device) |
| Voice matching for Me | Starter (free with the device) |
| Name Speakers, with segment-only or all-segments naming | Starter (free with the device) |
| Number of Speakers choice before generation | Starter (free with the device) |
| Voice matching for contacts | Premium or Elite |
| Contact Insights generated after you tag a contact | Premium or Elite |
Every Starter row stays in Premium and Elite as well.
Premium costs $14.99 a month or $119.99 billed yearly. Elite costs $39.99 a month or $359.99 billed yearly (US list prices, checked Oct 26, 2026).
How does contact matching work on Premium or Elite?
Contact matching starts with Tag Contact's Audio. You pick a clip of a contact's voice from a recording, and the app uses that clip to match the same voice in the note. When the match finishes, the speaker names in that note refresh.
Tag Contact's Audio has two entry points on a contact. When a new contact has no content yet, the option shows on the contact detail page. At any time you can also open the three-dot menu in the top right corner of the detail page and choose it from there.
The clip has to be long enough and clear enough. If it is too short or the sound quality is low, a message asks you to choose a different clip, and Save stays unavailable until you do. The app builds Contact Insights and a Profile from the note you tagged and leaves your other recordings alone.
Speaker recognition meeting notes: what does Starter give you for the other people in the room?
Starter gives you separation and naming. Every voice other than yours shows as its own speaker, and you name each one through Name Speakers. Once you name a voice in a recording, that name stays in that recording's transcript and note text. In the next meeting you name the voices again, or use Premium or Elite to have contacts matched by voice.
Which recording conditions affect speaker labels?
The actual performance of transcription, speaker identification and AI content is affected by recording distance, environmental noise, speaking style, network and language. Name Speakers is the step where you review and correct the labels.
The hardware decides how clean the audio is before the app starts labeling. Both devices run the same app and the same labeling steps.
| Comu Note Pro | Comu Action Pro | |
|---|---|---|
| Microphones | Two MEMS microphones plus one VPU | Six microphones |
| Indoor pickup distance | 5 m | 7 m |
| Continuous recording | 45 hours | Up to 70 hours |
| Storage | 16 GB | 16 GB |
| Speaker separation, Voice Profile, Name Speakers | Same app steps | Same app steps |
Comu Note Pro is 3.1 mm thick and weighs 27.6 g, with a 107-day standby. Comu Action Pro is built for settings from a one-to-one conversation to a whole meeting room.
Two steps apply to speaker labels:
- Set the Number of Speakers when you know it. Choose 1, 2 or 3+ instead of Auto when the headcount is clear.
- Name speakers after the transcript is ready. The Name Speakers dialog lists each voice found in the recording.
Where to get Comu Action Pro
Comu Action Pro costs $179.99 at store.comu.com (checked Nov 9, 2026), and Comu Note Pro costs $128.99. The free Starter plan comes with the device and includes Voice Profile, speaker separation and Name Speakers.
Frequently asked questions
Which AI voice recorder recognizes speakers by voice across meetings?
Comu Note Pro and Comu Action Pro recognize your voice across meetings once you set up a Voice Profile in the Comu app, and they label your turns as Me in new recordings. Matching other people by voice across meetings is a Premium or Elite feature (checked Nov 9, 2026).
How long does it take to set up a Voice Profile?
The recording takes 20 to 30 seconds. If the app already has microphone access, it skips straight to the recording screen.
Do I need the recorder to set up my Voice Profile?
No. The sample is recorded with your phone's microphone inside the Comu app, so the setup does not involve the device or its record button.
Can I delete my Voice Profile?
Yes. Open the Voice Profile screen, tap the three-dot menu, choose Delete and confirm. The entry returns to Not set. You can run the setup again at any time.
Does a Voice Profile change my old recordings?
A Voice Profile applies to new recordings. For an older recording, open it, tap Name Speakers and enter the names. The change updates that recording's transcript and note text.
How do I change "Speaker 1" to a real name?
Open the recording, tap Name Speakers, type the name next to Speaker 1 and save. The new name replaces the label in the transcript and in the note text. To change a single segment, tap its speaker label in the transcript and choose Apply to this segment only.
Is speaker identification included in the free plan?
Yes. Speaker separation, Voice Profile, voice matching for Me and Name Speakers come with the Starter plan that ships with the device. Voice matching for contacts and Contact Insights are Premium or Elite features.
Do I have to set up a Voice Profile to get speaker labels?
No. Speaker separation and Name Speakers work without a Voice Profile. The profile adds automatic Me labels in new recordings.
Where do I switch on speaker names for an exported transcript?
In the export window, check Include Speaker Names. It starts unchecked, and Include Timestamps starts checked.
Related reading
- How to Choose a Voice Recorder with Speaker Identification That Gets "Who Said What" Right
- Comu AI Transcription, Summaries & Feature Guide
- How AI Voice Recording and Transcription Work
- Comu Action Pro Price and Subscription Plans
- AI Notetaker for In Person Meetings (2026): 7 Options Compared by Price and Use Case
Sources
Pages were checked on Nov 9, 2026.
- Comu, Comu Action Pro
- Comu, Comu Note Pro
