Resources · Clone Your Voice
Claude now speaks in your voice
VoiceStudio is free, open source and runs on your own computer. A few seconds of recording is all it takes.
You want to make more videos. But every voiceover means: sit down, record, stumble, start over.
With VoiceStudio you record your voice once, for a few seconds. After that, Claude reads any text in your voice: voiceovers, explainer videos, podcasts, even audiobooks.
VoiceStudio is open source and runs entirely on your computer. It costs nothing. The project has more than 100,000 stars on GitHub and calls itself an alternative to ElevenLabs. According to the project it works in 646 languages.
The original lives here: github.com/debpalash/VoiceStudio
What you can do with it
- Voiceovers for reels and videos, without recording every sentence yourself
- Explainer videos: Claude builds the video, your voice goes on top
- Podcasts and audiobooks from your writing
- Other languages: your text in your voice, in another language
Important: only clone your own voice, or one you have permission to use.
What you need
- Claude Code. That’s the part of Claude that creates files and installs tools on its own. You’ll find it in the Claude desktop app (“Code” tab) or in the terminal (claude.com/claude-code).
- A paid Claude plan (Pro is enough).
- 10 to 15 seconds of clean recording of you, for example a voice memo.
No coding required.
Set up with one prompt
Paste this into Claude Code:
Install the VoiceStudio app on this computer and check that it runs. Follow this guide:
https://github.com/debpalash/VoiceStudio/blob/main/docs/install/agent.md
Then also install the skill: npx skills add debpalash/VoiceStudio
At the end, explain in plain English what you did and how I add my voice.
Claude does the rest. Depending on your computer it takes a few minutes, because the voice model has to download.
Clone your voice and use it
Put your recording in a folder and tell Claude:
Here's a recording of my voice: [path to the file].
Use it to create my voice clone in VoiceStudio.
Then read this text in my voice and save the file to my desktop:
"[your text]"
Listen to the result. If something sounds off, use a longer or calmer recording and try again.
What it can’t do
It won’t sound perfect on every sentence. Long, calm text works better than loud, emotional text. And some voice models aren’t licensed for commercial use. Check in VoiceStudio which license your model has before you use it for clients.
Who’s behind VoiceStudio
VoiceStudio is an open source project on GitHub (AGPL-3.0). This guide is mine. I’m Myrko Thum, a sales funnel expert with more than 550 funnel projects.
FAQ
Does it cost anything?
VoiceStudio is free and runs locally, no subscription. For Claude Code you need a paid Claude plan. Pro is enough.
Can I use the voices commercially?
The app itself, yes (AGPL-3.0). But every voice model has its own license, and some are for personal use only. VoiceStudio shows you the license in the Model Manager. Check it before you use a voice in client videos.
How good does it sound in my language?
Test it with your own text. Less common languages can be shaky, and in a second language your voice keeps your accent.
Do I need a fast computer?
A Mac with Apple silicon or a PC with a graphics card is ideal. It also works without one, just slower.