← Funkcje

Give an AI influencer your own voice

17 lis 2026 · 2 min czytania · ViralShaker team

Plany
Starter, Growth and Pro (any plan with AI influencers). Voiceovers cost 2 credits per started 1,000 characters.
Narzędzia dla agentów (MCP)
set_character_voiceremove_character_voicespeak_as_characterlist_voices

Every AI influencer comes with a choice of preset voices. If you want the face of your brand to sound like you, or like a team member who agreed to it, give them a real voice instead. Record once and every talking video, character swap and voiceover uses it.

How to set it up

  1. Open your AI influencer and choose to use a recorded voice.
  2. Record 10 seconds to 3 minutes of clear speech, or upload an audio file or a video of the person talking.
  3. Confirm the voice is yours, or that the speaker agreed.

That's it. Nothing else changes in how you make videos.

Where the voice is used

  • Talking videos. Talking head, product spokesperson, street interview and the other speaking formats are made as usual, then the speech is re-voiced in the recorded voice. Timing is kept, so lip sync still matches, and every clip sounds like the same person (see AI UGC videos).
  • Character Swap. "Match the voice" uses the recording instead of a preset.
  • Voiceovers. Turn up to 5,000 characters of text into speech in that voice, for 2 credits per started 1,000 characters.

Languages

The voice works in English, Spanish, French, Mandarin, Hindi, Arabic, Portuguese, German, Italian, Indonesian, Russian, Japanese, Vietnamese, Korean, Turkish, Polish, Dutch, Ukrainian, Thai, Swedish and Czech, with other languages detected from the text.

Getting a good recording

  • Record somewhere quiet, close to the microphone, without music underneath.
  • Talk the way the influencer will talk in posts: conversational, at your normal speed.
  • A minute or two of varied sentences works better than ten seconds of one line.
  • A video of you talking works too; only the sound is used.

Switching back

Remove the recorded voice at any time and the influencer goes back to its preset voice. The recording stays in your Media Bank.

For AI agents

set_character_voice takes a Media Bank recording (upload it first), speak_as_character makes a voiceover with a max_credits cap, and remove_character_voice switches back. See the MCP server.

Using a real voice is also a trust signal: read whether audiences trust AI influencers and the ethics of AI UGC.

Częste pytania

How much audio do I need?

Between 10 seconds and 3 minutes of clear speech. Record it in the app or upload an audio file or a video of you talking.

Does lip sync still match?

Yes. The video is made first, then the speech is re-voiced in your voice with the timing kept, so the lips still match.

Can I use someone else's voice?

Only with their agreement. You confirm the recording is your own voice, or that the speaker agreed to it being used.

Can I go back to a preset voice?

Yes, at any time. The recording stays in your Media Bank in case you want it again.

Powiązane artykuły