How to see who is talking in live captions

Captions that show words but not who said them leave you guessing at a table, in a meeting or at a family gathering. The fix is to use an app that separates speakers and lets you name them. Professional captioning standards ask for speaker identification, and the Described and Captioned Media Program's Captioning Key says establishing who is speaking, on screen or off, is vital for clarity¹.

This guide shows what the standard asks for, what the tools you can use today do, and how to set up Captio AI, a live captioning app for deaf and hard of hearing people, so names appear on its captions on iPhone.

Why do captions need to say who is talking?

Captions need to say who is talking because following a group is hard without it: the Government of Alberta's health pages note that everyone has trouble hearing in groups, especially when more than 1 person is talking at the same time². Words in a caption stream are only half of the conversation. In a group dinner or a team meeting, the other half is who said them and who to look at when you reply.

The Captioning Key, a standard for caption quality, sets the rules for professional captions [1]:

• Use caption placement first. Put the caption under the person who is talking when you can.
• Do not name a speaker until the person has been introduced in the audio or in on-screen text.
• When the name is known and placement does not work, put the name in parentheses on a line of its own.
• When the name is unknown, use the same information a hearing viewer has, such as "speaker #1".

A live captioning app on a phone cannot place captions under people at a table, so the name or label does that job.

Captio AI vs Pixel Recorder, Live Transcribe, Apple Live Captions and Teams: which shows who is speaking?

Captio AI, Google Pixel Recorder, Google Live Transcribe, Apple Live Captions and Microsoft Teams live captions all turn speech into text, but they differ on speaker names, platforms and languages. The matrix below compares them on the same rows, as seen in October 2026. "Not stated" means the maker's page we opened does not say, which is not the same as "No".

Speaker names matrix

The Captio AI cells come from the app's release notes and App Store listing⁵. Every other cell comes from the maker's own page, linked in the source list.

Which one fits which situation?

A Spanish and English family dinner is a case where language matters: Captio AI captions 60 languages, including speech that switches languages mid-sentence. The 1.1.3 release notes do not say which languages saved voices cover, so test with the people you talk to. Apple says some features may not be available in all regions or languages⁶.

How do I get names on captions with Captio AI?

Captio AI shows names on captions after you save the voice of the person, which takes about 20 seconds of that person talking. Here is the order that works:

1. Open Settings on iPhone, go to the Captions section and make sure Separate Speakers is on.
2. Open Recognize voices and add yourself first, then the person you talk to most.
3. Record that person talking, or choose a recording of them. The app needs about 20 seconds of actual speech.
4. Start Live Captions. Captions for that person now show their name instead of Speaker 2.

Voices stay on your phone. The app can also ask "Who's Speaker 2?" when it keeps hearing a voice it does not know, and each answer helps it recognize that person next time.

Which use cases do saved voices help?

Saved voices help most in settings where the same few people talk to you again and again. Each of these has a page on the site with the details.

• Group dinners and family gatherings. In group settings, the people around the table are usually the same, so saving 1 or 2 voices gives names on most lines.
• Team meetings. In meetings, a saved voice means a colleague's lines carry their name, and the names carry into the transcript and exports.
• Everyday conversations. In everyday conversations, saving your partner or a close friend means their words are named wherever you talk.
• Doctor's appointments. In a doctor's appointment, you can rename a speaker afterwards, for example to the doctor or the relative who came along.

On the free plan, Captio AI recognizes your own voice plus 1 other person, and Captio AI Plus allows more people.

What if I did not save a voice?

If you did not save a voice, Captio AI still separates speakers and labels them Speaker 1, Speaker 2 and so on, which matches the standard's advice to use a label when the name is unknown¹. You can tap a speaker's tag in a saved recording and say who it is.

What tips make speaker names more reliable?

Speaker names are easier to follow when people take turns, because everyone has trouble hearing in groups when more than 1 person talks at once². Alberta's health pages suggest asking that only 1 person talks at a time². Google's Live Transcribe help page advises holding the microphone near the person you want to capture⁴, and the same sense applies to any captioning app.

Where is Captio AI not the best choice?

Captio AI is not a replacement for a booked captioner when you are entitled to one, such as CART at a class or conference, which is covered in What is CART captioning?. Saved voices are on iPhone with version 1.1.3, and Android availability of this feature has not been confirmed. Live captions also need an internet connection. On Android, Google's Live Transcribe is a different option, though its help page does not describe speaker names⁴.