That open-source tool for making AI videos in ComfyUI just got a major overhaul, they really packed in the features.

Been following this free open-source AI video node that just dropped a major update—basically a full rewrite. The dev said they went full-time for a month and a half on it, and judging by the feature list, they really put in the work. Now you can do the whole video editing pipeline right inside the node: extend videos using prompts, keyframes, and audio combos, plus trim, split, and stitch clips directly on the timeline.

It supports IC-LoRA too—just drag your video onto the corresponding track and it’s good to go. For audio, there’s inpainting that seamlessly blends imported sounds with generated ones, and you can extend while feeding prompts.

There’s also a beta “reshoot” mode where you can select a segment and regenerate what happens inside—early experimental stuff. Timeline settings can be saved as JSON for reuse. Should be a big help for the open-source AI video community.

The best part is doing everything inside the nodes—no more exporting back and forth.

The re-shoot mode sounds interesting, but is the beta version stable or not?

IC-LoRA just drag the node in and it works, super user-friendly.

Open source, free, and this comprehensive — the dev really has passion for it.

The timeline can save as JSON, so you can keep editing without starting over.

Audio inpainting? That’s a rare feature, looking forward to seeing it tested.

Bookmarked.