Offline Windows tool that clones your face and voice from a video, then makes avatar videos from text or audio.
Make avatar videos from text or audio on a Windows PC, with no internet connection.
Duix.Avatar is a fully offline video synthesis tool from DUIX.COM that clones your appearance and voice, then drives the avatar from text or audio. It suits people with a gaming-class Windows PC who want presenter videos without a subscription.
The README says it runs on Windows 10 or Ubuntu 22.04 through Docker, using bundled speech-recognition (fun-asr) and voice-cloning (fish-speech) services. You supply a video of yourself and it produces new videos of that avatar speaking your text.
Suits individuals and small teams who want a Synthesia-style presenter they own. It does not suit you if you need real-time conversation (the README says it is non-real-time) or a machine without a strong NVIDIA GPU.
Builds an avatar from a video of a real person.
A bundled voice-cloning service speaks your text in a cloned voice.
Drive the avatar with typed text or an audio file.
The README says no internet connection is needed after setup.
Windows installer or Docker Compose, with a lite compose file.
Tagged releases are published on GitHub.
Free to run under the DUIX.COM Community License for products with up to 1,000 monthly active users. Beyond that you request a commercial licence from DUIX.COM. The README points to duix.com for hosted real-time interaction; no price is listed there. The cost is your own hardware.
767 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
Only below 1,000 monthly active users under its community licence. Above that you must request a commercial licence from DUIX.COM, and you must show "Built with DUIX.COM" attribution.
Yes. The README describes it as a fully offline tool once the Docker images and models are downloaded.
No. The README says it does non-real-time synthesis and points to duix.com for interactive avatars.
See Zekai first in Google