Does it cost anything?

Connecting is free. Generations use the credits in your Voices AI account at the same rates as the Voices AI apps: speech costs 1 credit per character (0.5 with turbo), changing the voice in a recording or removing its background noise costs 1,000 credits per minute of audio, a transcript costs 100 credits per minute, and a song costs 1,000 credits. New Voices AI accounts include free starter credits.

Which voices can I use?

The MCP server includes the Voices AI voice library in multiple languages and the voices you cloned in the Voices AI apps, which only you can use. Your assistant can search them by name, language or category and play samples before generating. To clone a new voice, use Voices AI Studio or the mobile apps.

Where do my generations go?

Your assistant shares a link to each MP3 in the chat. Every generation is also saved to your Voices AI account history, so you can find it again in Voices AI Studio and the apps.

What can the connector access?

Only your Voices AI account: your credit balance, subscription status and generation history, plus any recordings you choose to share for voice changing. It spends credits only on the generations your assistant creates for you, and it does not receive your conversations.

Are songs I create private?

Songs appear on the public Voices AI song pages by default, as in the apps. Subscribers can ask for a song to be created as private.

How do I disconnect Voices AI?

Remove the Voices AI connector in your AI assistant's settings. Your assistant can no longer use your account after that, and your existing generations stay in your Voices AI account.

How do I change the voice in my own recording?

Ask your assistant to change the voice and tell it which voice to use. In ChatGPT you can attach the audio or video file directly. You can also share a direct link to the file, or your assistant gives you a secure upload page on getvoices.ai. Recordings up to 60 minutes work, and you get the new voice back as an MP3.

Voices AI in Claude & ChatGPT Ask for audio, get it in the chat

Connect once with your Voices AI account, then ask your assistant for a voiceover, a new voice on your recording, a song or a transcript.

Add to Claude

Also works in ChatGPT, Cursor, VS Code and other MCP apps. See setup

Read our welcome message in a warm, friendly voice.

Voices AIgenerate_speech

Here's your voiceover, read by Hazel.

Hazel“Welcome to Voices AI. This is my voice and I can speak whatever you want. Just type in your text and generate an ultra-realistic audio. Try now voices AI yourself.”

0:11

Works with: Claude, Claude Code, ChatGPT, Cursor and other MCP clients
Sign-in: Secure OAuth with your Voices AI account
Voices: The voice library in multiple languages, plus your own cloned voices
Pricing: Uses your Voices AI credits at the same rates as the apps
Speech: Up to 10,000 characters per request, MP3 output
Voice changer: Audio or video up to 60 minutes, returned as MP3
Songs: From a short description or your own lyrics
Transcripts: Text with timestamps or SRT and VTT subtitles, from audio or video

Connect in a minute

Each client asks you to sign in with your Voices AI account once. After that, the tools are available in every chat.

Server URL

https://mcp.getvoices.ai/mcp

Claude

Add to Claude

Or add it yourself:

1Open Customize → Connectors and choose Add custom connector.
2Enter the name Voices AI and the server URL, then select Add.
3Select Connect, sign in on getvoices.ai and allow access.
4In a chat, turn on Voices AI in the tools menu and ask for audio.

ChatGPT

Add Voices AI as a custom connector with the server URL using developer mode in ChatGPT's settings, then sign in with your Voices AI account. Developer mode availability depends on your ChatGPT plan.

Claude Code

Add the server from your terminal, then run /mcp in Claude Code and choose Authenticate.

claude mcp add --transport http voices-ai https://mcp.getvoices.ai/mcp

Cursor, VS Code and other clients

Any client that supports remote MCP servers with OAuth can connect. Add the server URL to its MCP configuration, for example:

{
  "mcpServers": {
    "voices-ai": {
      "url": "https://mcp.getvoices.ai/mcp"
    }
  }
}

Give your agents a voice

Agents like Claude Code, Codex and Cursor can use Voices AI while they build. They pick a voice, generate the audio and save the file into your project.

Add a short spoken intro to the landing page hero, in a warm female voice from Voices AI.

voices-ai · search_voicescategory: "professional", language: "en"
voices-ai · generate_speechvoice_id: "hazel"
Bashcurl -o public/audio/intro.mp3 "https://cdn3.getvoices.ai/audio/…"
Updatesrc/components/Hero.tsx+6 −1

Done. The hero now opens with a short intro read by Hazel.

Websites and apps
Narration, onboarding audio and spoken product tours, saved straight into the repo.
Videos
Voiceovers, songs and music for videos your agent edits or renders in code.
Games
Voice lines for every character, in many languages, ready to drop into the build.
Podcasts and courses
Scripts, docs and lessons turned into narrated episodes, with transcripts for the show notes.

What your assistant can do

The Voices AI MCP server gives your assistant a set of tools. It decides which ones to use based on what you ask.

Voices AI

https://mcp.getvoices.ai/mcp

search_voices
Find voices in the Voices AI library and your own cloned voices by name, language or category, with a portrait and a sample to preview.
generate_speech
Uses credits
Turn text into natural speech and get a link to the MP3. Choose high quality, or turbo for faster, cheaper results.
change_voice
Uses credits
Replace the voice in an audio or video recording and get the new audio back as an MP3. Words, timing and emotion stay the same.
isolate_voice
Uses credits
Remove background noise, music and room echo from an audio or video recording so only the voice remains.
transcribe
Uses credits
Turn the speech in an audio or video recording into text, with timestamps or as SRT and VTT subtitles.
create_song
Uses credits
Create an original song from a short description or your own lyrics and a music style. Each request produces two versions.
get_generation
Check a speech generation, voice change or cleanup and get the audio when it is ready.
get_transcription
Check a transcription and get the transcript when it is ready, page by page for long recordings.
get_song
Check a song while it generates and get the audio and cover art when it is ready.
create_upload_link
Create a secure upload page for a recording you can't attach or link, so your assistant can change, clean up or transcribe it.
list_history
Show your most recent speech, voice changes, songs, transcripts and cleanups, with links.
get_account
Check your credit balance and subscription status.

Things to ask

Write the way you normally would. Your assistant picks a voice, writes the text if needed and creates the audio.

Find a German voice on Voices AI and turn this paragraph into a voiceover.
Narrate this blog post as an MP3 with a calm professional voice.
Write and create an upbeat pop song about our product launch.
Make a short voice message wishing my team a great weekend.
Show me links to my last three Voices AI generations.
Change the voice in this video to another Voices AI voice.

Frequently asked questions

MCP (Model Context Protocol) is an open standard that lets AI assistants use tools from other services. The Voices AI MCP server lets assistants such as Claude and ChatGPT create speech and songs, change or clean up the voice in recordings and transcribe them with your Voices AI account, so you can work with audio without leaving the chat.

Want more voices and features?

Voices AI Studio, the mobile apps and Voices AI Live add voice cloning, real-time voice changing and more.

Open Voices AI Studio See pricing

Voices AI in Claude & ChatGPT Ask for audio, get it in the chat

Connect in a minute

Claude

ChatGPT

Claude Code

Cursor, VS Code and other clients

Give your agents a voice

Websites and apps

Videos

Games

Podcasts and courses

What your assistant can do

search_voices

generate_speech

change_voice

isolate_voice

transcribe

create_song

get_generation

get_transcription

get_song

create_upload_link

list_history

get_account