• AUDIO
  • DATA
  • HUMAN
  • LOGS

How to change voice identity with your AI

Connect Resemble AI to your agent to swap voices in existing recordings. Once connected, you can hand your AI any audio file and tell it which voice to use instead.

Ask AI about this page

Short answer

How can I change the voice in an existing recording?

You provide an audio file and a target voice profile to your agent. The AI uses speech-to-speech processing to map the new voice onto the original performance, preserving the original delivery and inflection. You'll get a new audio file with the updated identity.

The outcomes

What you get from a voice swap

The AI processes your request into specific results.

AUDIO

New voice files

The AI generates a recording that matches the original timing but uses the target voice. You get a clean file ready for use.

DATA

Extracted clips

If you only need a specific segment, the AI can isolate that portion of the audio. This gives you a standalone clip.

HUMAN

Voice profile setup

The AI cannot invent a voice from nothing. You must provide the target voice identity or a profile for the AI to use.

LOGS

Processing history

The AI keeps track of the clips it has processed. You can ask your agent to find or reference previous conversions.

The workflow

What your AI does when the audio arrives.

Your agent handles the audio manipulation through these specific steps.

  1. Isolate the audio

    The AI identifies the specific audio segment you want to work with from your source.

    get_clip

  2. Perform the swap

    The AI maps the target voice onto the original audio, maintaining the original cadence and emotion.

    speech_to_speech

  3. Verify the clip

    The AI checks the processed segment to ensure it matches the requested parameters.

    get_clip

  4. Finalize the file

    The AI prepares the converted audio for you to use in your project.

    speech_to_speech

Try it

Copy these to start.

Use these prompts to trigger the voice conversion process.

Starting points

These are opening lines. Swap the bracketed text for your actual file names or voice IDs.

Resemble AI Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_lELKY7eitB2gZRvJGRTlTOUHxzdfjT5MrnBpOAJW/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Resemble AI capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "resemble-ai-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_lELKY7eitB2gZRvJGRTlTOUHxzdfjT5MrnBpOAJW/mcp"
    }
  }
}
Copy into chat04

  • Use speech_to_speech to change the voice in my recording 'interview_01.wav' to the 'Professional Narrator' profile.

  • Get the clip from 'podcast_segment.mp3' and change the identity to 'Deep Bass Voice'.

  • I need to swap the voice in 'customer_feedback.wav' using the 'Friendly Assistant' profile.

  • Take the audio from 'demo_clip.wav' and apply the 'Calm Female' voice identity.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Resemble AI once, then ask.

One link is all it takes. Your login stays encrypted on our side, and you can start working immediately.

Connect Resemble AI to your AI

FAQ

How the voice swap behaves

  • 01

    Will the new voice sound like the original person?

    No. The goal is to change the identity. The new voice will sound like the target profile you selected, not the original speaker.

  • 02

    Does the AI keep the original emotion and speed?

    Yes. The process is designed to preserve the original timing, inflection, and emotional delivery.

  • 03

    Can the AI create a brand new voice from a text prompt?

    No. The AI requires a target voice profile or an existing voice identity to perform the swap.

  • 04

    Can I edit the words in the audio?

    No. This tool changes the voice identity of existing audio. It does not perform text-to-speech or edit the spoken script.

  • 05

    Can the AI delete my original audio files?

    No. The AI only reads and processes the clips you provide; it does not have the capability to delete your source files.

  • More questions about Resemble AI? The Connector page answers them. See everything the Resemble AI Connector can do

Connect Resemble AI to Claude, Cursor, ChatGPT & more