vlo 0.2.0: ComfyUI-Powered Editor Gives Fine Control Over AI Video Generations
This open-source tool lets you salvage near-perfect clips with inpainting and motion cues.
Get AI news that actually matters
One email a day. Zero fluff. Join 10,000+ professionals.
Developer PxTicks has released vlo 0.2.0, an AI video editor built on ComfyUI that aims to give users more control over AI-generated video than typical tools. The project focuses on reducing the frustration of imperfect generations by providing tools to fix, correct, and enhance clips rather than re-rolling the dice. It works with standard ComfyUI workflows but allows augmentation through special rules files that enable the workflow to read masks and motion cues directly from the timeline. This design philosophy sets vlo apart from other AI video editors that treat generation as a one-shot process.
The editor supports video-to-video tooling for inpainting, correction, foley, and creative effects. It handles technical mismatches between different video models—for instance, the different permitted aspect ratios of Wan and LTX—so users can combine the best outputs from each. vlo also offers a layered editing system, allowing complex edits without constant jumping between ComfyUI and a separate video editor. The developer provided a demo video made entirely in vlo using Wan and LTX (plus two images from nano banana). The project is open-source on GitHub, and a Runpod template is available for easy deployment.
- Designed for control: reduces rerolls by providing inpainting, correction, and foley tools for almost-perfect generations.
- Works with generic ComfyUI workflows plus special rules to read masks and motion cues from the timeline.
- Smooths technical mismatches between models like Wan and LTX, handling differing aspect ratios and enabling hybrid workflows.
Why It Matters
Gives AI video creators granular control to fix and enhance generations, bridging gaps between major video models.