Free public beta · Windows
Your voice.
Your edit.
Your creative twin.
CueTwin is a desktop video editor you can direct by typing or speaking. A built-in local assistant edits the timeline, local voice cloning narrates your script in your own voice, and the final say stays on your timeline.
Free during the public beta. Windows 10/11 64-bit.
See it in action
What it does
From raw take to finished edit.
Edit by typing or speaking
Tell the built-in assistant what you want: cut dead air, split, add titles, change the canvas, export. It runs on local models on your PC, so no API key is needed.
Narrate in your own voice
Voice Studio makes an instant clone from your recorded samples, or trains a professional clone locally. Recordings and training stay on your PC.
A real multi-track timeline
Media library, preview, and a multi-track timeline. Move, trim, split, add text and transitions, with undo and redo.
Trim silence and dead air
Cut dead air, leftover hiss, and clicks, then close the gaps. Or ask the assistant to remove silence across a long recording.
Smart edit for screen recordings
Keeps the part of the screen that matters, speeds up stretches where an AI agent is only writing, and cuts time spent in unrelated apps.
Record and export
Record a display or a single window (games included) with system audio. Export through FFmpeg, using a hardware encoder when one is available.
The beta
Free to try. Help shape it.
- Free during the public beta, with exports up to 1080p and instant voice cloning.
- Higher-resolution export and professional voice training will be part of a paid Creator licence.
- Paid licences come later through Lemon Squeezy. Pricing hasn’t been decided.
- Found a problem? Report it on GitHub Issues. Every report helps.
Before you download: the installer itself is kept small enough for GitHub Releases. On first launch, CueTwin downloads its local AI models (roughly 8 GB), so you need an internet connection and free disk space for that. After that, the assistant runs locally.
Voice cloning also needs Python 3.11 (64-bit) and an internet connection to install the voice runtime and download the speech models. See the system requirements.