How to clone your voice in Google Flowthrough your avatar, not a voice cloner
Google Flow has no tool that clones only your voice. Your voice gets into Flow through your avatar, which Google says features your own visual and audio likeness: you record your face and voice on your phone, then type @me in a prompt. It is not available in the EEA, the UK or Switzerland. For your voice reading an exact script in any of 90+ languages, as an MP3 you can lay over a Flow clip, clone it in a voice tool such as oVideo Voice Clone.
Updated
Short answer
Can you clone your voice in Google Flow?
Only as part of your Flow avatar. Google’s help page says a Flow avatar features your own highly accurate visual and audio likeness; you start it on a computer, record your selfie and voice on your phone, and use it by typing @me in the prompt box. Flow’s “custom voices” are something else: a preset voice restyled from a description, not a clone. Flow also does not dub or change what people say in uploaded videos, in any region. If you need your own voice reading your exact words, clone it in a voice tool and add the MP3 to the Flow video in an editor.
- Your voice in Flow: the avatar, used with @me
- Custom voices: restyled preset voices, not clones
- No avatar in the EEA, UK or Switzerland
- No dubbing of uploaded videos in Flow
Read on support.google.com/flow and support.google.com/accounts on November 20, 2026.
Way 1, inside Flow
Create your Flow avatar with your voice
The steps as Google’s Flow help page gives them on November 20, 2026. Google’s tips: hold the phone at eye level, record in a quiet place with no other voices in the background, and only the account owner records; minors must not record themselves.
- 01
Open Flow on a computer
Go to Google Flow and open an existing project or start a new one.
- 02
Create avatar
At the top right, select your profile picture or initial, then Create avatar, then Get started.
- 03
Finish on your phone
Scan the QR code with your phone or tablet and follow the on screen instructions to record your face and voice.
- 04
Use @me in a prompt
In the prompt box, type @me, describe how your avatar should appear, then select Generate.
- 05
Delete it any time
Profile picture, Manage avatar, Delete avatar; or in your Google Account under Recordings for your avatar.
Flow and voices, question by question
What Google’s pages say, and what they leave out
What are Flow custom voices?
On Omni Flash: Add, Voices, Create New Voice. Pick a preset Base Voice, name it, describe the Voice Performance (Google’s example: slightly raspy with a New York accent), preview 8 seconds, then Sync and Save. A restyled preset, not your voice.
How do voice references work?
With Omni Flash and Ingredients, add a single speaker voice from Voices and refer to it as @Voice in the prompt to keep a character’s voice across clips. Voice references only work on generations that use ingredients.
Can Flow change what someone says in my video?
No. Google lists editing or generating what people say in uploaded videos, adding dialogue or dubbing their voice, as unsupported in every region. You can still edit visuals, background audio, music and sound effects.
Can the avatar read my exact script?
Not stated. Google’s pages do not say how dialogue for the avatar is written, which languages it speaks in Flow or whether you can export audio only.
Is the avatar shared with others?
In Flow, avatar data cannot be shared through Flow Tools or public project links. The same avatar works in Gemini, Flow and Vids.
Who can use it?
Flow is 18 and over. Avatars are not available in the EEA, the UK or Switzerland; Google’s Gemini avatar page says avatars are only supported in English for now.
Read on Google’s Flow, Gemini and Google Account help pages on November 20, 2026.
Way 2, the workaround
Clone your voice, then add the MP3 to your Flow video
Use this when you are in Europe or the UK, when you need your exact words, another language, or a voiceover on many clips. It works with oVideo Voice Clone or ElevenLabs; the details here are oVideo’s.
- 01
Make the Flow clip without dialogue
Prompt the scene with “no dialogue” or “ambient sound only” so a voiceover will not clash, then download the video from Flow.
- 02
Clone your voice
In oVideo Voice Clone, record in the browser or upload a file, even a video of you talking (MP4, MOV, up to 100 MB). Tick that it is your voice. Ready in about a minute.
- 03
Type the voiceover
Up to 5,000 characters, written in the language you want to hear, from 90+. The credits show before you click: 25 per started 30 seconds.
- 04
Download the MP3
Your last 20 voiceovers stay listed in the studio.
- 05
Put them together
Drop the Flow clip and the MP3 on one timeline in CapCut, iMovie, Premiere or Clipchamp, line them up and export.
Flow prompt
“Slow drone shot over a misty pine forest at sunrise, soft wind, birdsong, no people, no dialogue.”
Your voiceover
“I used to think mornings were for rushing. Then I came up here, and the forest taught me to wait.”
Flow prompt
“Close up of hands kneading bread dough on a floured wooden table, warm kitchen light, no dialogue.”
Your voiceover
“My grandmother never measured anything. A handful of this, a pinch of that, and the bread was always perfect.”
Flow prompt
“A small red tram climbing a steep street at golden hour, city sounds only, no dialogue.”
Your voiceover
“Day three in the city. I stopped checking the map and just followed the tram.”
Compared, November 20, 2026
Your voice in Flow, or next to it
What each option does by its own pages, read on November 20, 2026. Not stated means Google’s pages did not say. oVideo is our product, so read its column with that in mind.
| Feature | oVideo Voice Clone (ours) | Flow avatar | Flow custom voice | Gemini API voice replication | ElevenLabs |
|---|---|---|---|---|---|
| Is it your own voice? | Yes, cloned from your sample | Yes: “your own highly accurate visual and audio likeness” | No: a preset voice restyled by a description | Yes, from a 10 to 30 s clip plus a consent statement | Yes, instant or professional clone |
| What you give | A browser recording, or an audio or video file up to 100 MB | A selfie and voice recording on your phone | A text description | Two recordings of the same adult speaker | A sample of 1 to 2 min for an instant clone |
| What you get | An MP3 of any text you type, up to 5,000 characters | Flow videos with you in them, via @me | A voice for Flow videos made with Ingredients | Speech in your own app | Audio files, dubbing, API |
| Exact script, audio only file | Yes | Not stated | Not stated | Yes, through the API | Yes |
| Languages | 90+ (Eleven v4, per ElevenLabs) | Avatars in English only, per Google’s Gemini avatar page | Not stated | Consent statement in supported languages | 90+ with Eleven v4 |
| Where it works | Browser on a phone or a computer; no regional block stated | Not in the EEA, UK or Switzerland | Some customizations only in certain regions | Paid tier API projects, including the EEA and UK | Web and API |
Sources: support.google.com/flow answers 17102997, 16353334, 16352836 and 16353544; support.google.com/accounts answer 17100665; support.google.com/gemini answer 16984474; aistudio.google.com/docs/voice-replication; elevenlabs.io/pricing.
Which one should you pick?
Pick by the video you are making
- 1You on camera in a Flow scene, in English, outside Europe and the UK: the Flow avatar with @me.
- 2A character voice that stays the same across Flow clips: a Flow voice reference or custom voice.
- 3Your voice narrating Flow clips, word for word: oVideo Voice Clone or ElevenLabs, then an editor.
- 4Your voice in Spanish, Turkish, Hindi or another of 90+ languages: oVideo Voice Clone; type the text in that language.
- 5A voice clone inside your own app: the Gemini API voice replication or ElevenLabs.
- 6A stock AI voice on a clip, no editor at all: oVideo Add AI Voice, with captions in the same step.
Checklist
What to check before you voice a Flow video, answered for oVideo
Does it work in Europe and the UK?
Voice Clone runs in the browser and has no regional block like the Flow avatar’s.
Does it read my exact script?
Yes, word for word, up to 5,000 characters per generation.
Does it lip sync to the Flow clip?
No. It makes an MP3; use it as narration over clips where nobody talks to camera.
Which languages?
90+, the languages ElevenLabs Eleven v4 speaks.
Price?
A voice is 50 credits once; speech is 25 credits per started 30 seconds; Starter is $10 a month for 400 credits. Failures are refunded.
Free?
No. Free accounts can record; creating a voice needs a plan.
oVideo facts checked in the app on November 20, 2026.
Google AI Studio
Can Google AI Studio clone my voice?
Yes, as a developer feature. Google AI Studio’s Voice Replication page (read November 20, 2026) lets you record or upload two clips in the browser: a 10 to 30 second clip of clean, natural speech, and a clip of the same adult speaker reading Google’s consent statement, “I am the owner of this voice and I consent to Google using this voice to create a synthetic voice model.” You preview the voice and copy a voice ID into your app.
It needs a paid tier Gemini API project, and unlike the Flow avatar it is available in the Gemini API regions that include the EEA and the UK. Stored voices are kept for a year from their last use, up to 200 per project, and deleting one removes the profile. Every result carries an inaudible SynthID watermark, and recordings that are synthetic or already watermarked are rejected.
It is a tool for building apps, not a voiceover editor: you write code or use the playground to make speech. If you just want your voice reading a script as an MP3 in 90+ languages, a voice cloning tool such as oVideo Voice Clone or ElevenLabs is the shorter route.
Honest limits
When Flow’s own avatar is the better route
If you want to appear on camera and speak in the same generated scene, the Flow avatar does both at once, and oVideo Voice Clone does not: it makes audio, not video, and does not lip sync.
oVideo’s cloned voice also stays in the Voice Clone studio for now: it is not available inside oVideo’s own Add AI Voice, Story to Video or Shorts yet, so an editor joins the MP3 and the video.
Clone only your own voice, or one you have permission to use.
Commercial use of the audio you make on oVideo is allowed under the Terms, provided you had the right to use the voice you cloned.
Answers
Frequently asked questions
- Can Google Flow clone my voice?
- Through your avatar, yes. Google’s Flow help page (read November 20, 2026) says a Flow avatar features “your own highly accurate visual and audio likeness”, made from a selfie and voice recording on your phone and used by typing @me in a prompt. Flow has no separate tool that clones only your voice, and its “custom voices” restyle a preset voice from a description.
- How do I clone my voice in Google Flow?
- On a computer, open Flow, select your profile picture, then Create avatar and Get started. Scan the QR code with your phone and follow the steps to record your face and voice. Then type @me in the prompt box, describe the scene and select Generate.
- Can I add my own voice in Google Flow?
- Yes, through your Flow avatar, which Google says carries your own audio likeness; you record it on your phone and call it with @me. Uploading your own recorded voiceover file into a Flow video is not described on Google’s Flow help pages, so add that MP3 in a video editor.
- Does Google Flow do voice over?
- For generated clips, yes in a sense: Flow’s models make video with sound, and you can pick preset or custom voices for characters with Ingredients. For uploaded videos, no: Google lists adding new dialogue or dubbing voices in uploaded videos as unsupported in every region.
- Is the Flow avatar available in my country?
- Not in the EEA, the UK or Switzerland, by Google’s Flow help page. Flow itself is for people 18 and over, and Google’s Gemini avatar page says avatars are currently only supported in English.
- What are custom voices in Google Flow?
- A preset voice you restyle with a description, for example slightly raspy with a New York accent. You pick a Base Voice, describe the Voice Performance, preview 8 seconds and save it. It is not a clone of your voice.
- Can Flow dub my uploaded video in my voice?
- No. Google’s Flow page says that in every region it does not support editing or generating what people say in uploaded videos, including adding new dialogue or dubbing their voice. Add a voiceover afterwards in a video editor instead.
- Can I clone my voice in Google AI Studio?
- Yes, for developers. Google AI Studio and the Gemini API replicate a voice from a 10 to 30 second clip plus a recorded consent statement from the same adult speaker, on a paid tier API project. Every result carries a SynthID watermark.
- Is cloning my voice in Google Flow free?
- Flow’s avatar page does not state a price. Google’s Gemini avatar page says you need a Google AI plan to make a video with your avatar, and Flow says features vary by subscription tier. Check inside Flow before you plan around it.
- How do I add my own voice to a Google Flow video?
- Download the Flow clip, clone your voice in a voice tool such as oVideo Voice Clone or ElevenLabs, type the script, download the MP3, and put both on one timeline in CapCut, iMovie, Premiere or Clipchamp. Generate the Flow clip without dialogue so the voiceover does not clash.
- How do I delete my Flow avatar and voice?
- In Flow: profile picture, Manage avatar, Delete avatar. In your Google Account: Data and privacy, Other settings, Recordings for your avatar, Delete. Google says unused recordings are deleted after 3 years and videos you already made are not deleted.
- Does Google train its AI on my avatar voice?
- Google’s avatar page says your recordings are not used to train the generative models behind its services, except to address abuse or harm, and are used to check you are a real person giving permission.
- How do I clone my voice in Google AI Studio?
- Open the interactive Voice Replication experience on Google AI Studio’s speech generation page, record or upload a 10 to 30 second clip of yourself talking, then record yourself reading Google’s consent statement. Preview the voice and copy its voice ID to use with the Gemini text to speech models. It needs a paid tier Gemini API project.
- Is voice cloning in Google AI Studio free?
- Google’s Voice Replication page says it requires a paid tier (Tier 1 or higher) Gemini API project.
- Does Google AI Studio voice cloning work in Europe and the UK?
- Yes, per Google’s Voice Replication page: it is available in the supported Gemini API regions, which include the EEA and the UK. That differs from the Flow and Gemini avatar, which Google does not offer in the EEA, the UK or Switzerland.
- Can Gemini clone my voice?
- Two ways, by Google’s own pages: the Gemini app’s avatar carries your voice and likeness in English (not in the EEA, the UK or Switzerland), and the Gemini API replicates a voice for developers from a 10 to 30 second clip plus a consent recording. More on “Can ChatGPT clone my voice?”, which also covers Gemini.
- Which is easier for a voiceover in my own voice: Flow, AI Studio or a voice cloning tool?
- For narration over clips, a voice cloning tool: record once, type the script, download the MP3. Flow is easier when you want to appear on camera as your avatar, and AI Studio when you are building an app.
Related
- Clone my voice with AISteps, price and 90+ languages.
- Clone your voice for YouTubeFaceless videos in your own voice.
- My voice in another languageHear yourself speak Spanish and more.
- Best AI voice cloning appsTen apps and the job each one wins.
- Can ChatGPT clone my voice?And Gemini, Grok and Canva.
- Can CapCut clone my voice?An editor with a clone built in.
- PlayHT alternativesMove a cloned voice to a new tool.
- Bedtime stories in your own voiceStory texts to paste, MP3 routine.
- Is Veo 3 free?What Google’s video model costs.
- How to add AI voiceover to a videoPut narration on any clip.
- Add AI voice to videoA stock ElevenLabs voice on your own clip.
- Best AI voiceover for videoVoiceover tools compared.
- Add AI voice on iPhoneIn Safari, no app.
- Faceless YouTube channel with AINarrated videos without your face.
- Can AI videos be monetized on YouTube?YouTube’s rules on AI content.
- Eleven v4 voice cloneThe ElevenLabs model behind oVideo Voice Clone.
Sources, checked November 20, 2026
- Google Flow Help: Create & use your avatar
- Google Flow Help: Create videos (voice references, custom voices)
- Google Flow Help: Models & supported features
- Google Flow Help: Where you can use Google Flow
- Google Account Help: Create, use & manage your avatar
- Gemini Apps Help: Create and use an avatar
- Google AI Studio: Voice replication
Your own voice over your Flow clips
Record or upload a clip of you, tick that it is your voice, type the voiceover in any of 90+ languages and download the MP3.