VoicerAI
Voice cloning, voice design and dubbing on your own computer, connected to your ThinkFlyFlow workspace.
VoicerAI brings studio-grade voice tools to your desktop: clone a voice from a short recording, design voices from a description, turn scripts into finished audio, and upload what you make to ThinkFlyFlow without leaving the app. The cloning engine runs locally, so your recordings never have to leave your machine.
Windows 10/11 and macOS. A free ThinkFlyFlow account is required.
What you get
- Voice cloning from a reference recording
- Voice design from a written description
- Text to speech, stories and audiobooks
- Dubbing and transcription
- Local generation on CPU or GPU
- Direct upload to your ThinkFlyFlow workspace
Download VoicerAI
Install it, sign in with ThinkFlyFlow, and the studio opens.
Windows
A standard installer for Windows 10 and 11 on 64-bit PCs. An NVIDIA graphics card speeds generation up; it also runs on the CPU.
Download for WindowsmacOS
A disk image for both Apple silicon and Intel Macs. Apple silicon uses the Mac graphics chip, so it is the fastest option.
Windows installers are code-signed by SignPath Foundation, the free code signing programme for open source projects, so Windows does not warn about an unknown publisher. macOS builds are not yet notarised by Apple, so Gatekeeper may ask you to confirm the first launch. On macOS, allow VoicerAI under System Settings, Privacy and Security. Checksums for every release are on the releases page.
What it does
One studio for everything voice, running on your own hardware.
Clone a voice from a short recording
Give it a clean reference clip and the app learns the voice on the spot. No training run, no waiting days for a checkpoint.
Design a voice that does not exist
Describe age, tone and character to build a brand-new voice, then reuse it across every project in the studio.
Turn text into finished audio
Stories, audiobooks and long scripts, with pause and pacing control, exported as clean files you can edit elsewhere.
Dub and transcribe
Match a voice to an existing video, or pull a transcript out of a recording you already have.
Hundreds of languages
The engine behind the app covers 646 languages, so one cloned voice can speak to every market you sell in.
Runs on your own computer
The model runs locally on your CPU or GPU. Your recordings are not uploaded anywhere unless you choose to save them.
How it works
Create your ThinkFlyFlow account
The app is for ThinkFlyFlow members. Signing up is free and takes a minute; you can do it from the app itself when it first opens.
Download and install
Pick Windows or macOS, run the installer, and let the app set up its voice engine the first time it starts.
Sign in and start cloning
VoicerAI opens your browser, you approve the sign-in, and the app unlocks. Record or add a reference clip and generate your first voice.
Save your work to ThinkFlyFlow
Any voice file the app produces can be uploaded straight into your workspace, where you can organize it like everything else.
Your ThinkFlyFlow account opens it
The app signs you in through your browser. No password is ever typed into the desktop app itself.
Local by default
Recordings stay on your disk. Upload only the files you want backed up in your workspace.
Voice files up to 5 MB
Uploads are checked in the app and again on our servers, so a file that is too large is caught before it travels.
Questions
Do I need an account to use the app?
Yes. VoicerAI is a ThinkFlyFlow product, and it unlocks with your ThinkFlyFlow account. If you do not have one yet, the app opens the sign-up page in your browser for you.
Where are my recordings stored?
On your own computer. Voice cloning happens locally on your machine. Nothing leaves your device unless you upload a file to ThinkFlyFlow yourself.
How big can an uploaded voice file be?
Voice files uploaded to ThinkFlyFlow are capped at 5 MB each. The app checks the size before it uploads and tells you if a file is too large.
Does it work without a graphics card?
Yes. It runs on the CPU, just more slowly. An NVIDIA graphics card on Windows or an Apple silicon Mac is noticeably faster.
Do I need to be online to use it?
You need to be online to sign in. After that, generation runs locally. Being offline only stops the cloud upload and the initial engine download.
Is it signed software?
Windows installers are code-signed by SignPath Foundation, the certificate provider we use for open source builds, so Windows no longer warns about an unknown publisher. macOS builds are not yet notarised by Apple, so Gatekeeper still asks you to confirm the first launch. Allow it under System Settings, Privacy and Security.
Who signs the Windows installer?
SignPath Foundation. VoicerAI is open source under AGPL-3.0, and SignPath Foundation issues a free code signing certificate to open source projects so that users can verify a build really came from this repository. The signature is issued in their name, which is how the programme works.
Get VoicerAI
Download for Windows or macOS, then sign in with ThinkFlyFlow.