ID-V2V is an open video-to-video model released by Eyeline Labs, Netflix's research lab. It solves a specific problem: changing the look of a video without turning the person being filmed into someone else halfway through.
The workflow is different from what you'd expect from a video generator. You don't describe the scene in text. You upload the original video and edit a single keyframe with the look you want. The model treats that frame as the reference and propagates the change to all the others — keeping the face, expressions, and movement identical to the original clip.
Why this matters
- It's the real bottleneck for anyone making video with AI: generating is easy, keeping the same person in every frame is what's hard.
- Editing a frame is a job anyone with Photoshop can do — much more controllable than hoping a prompt gets it right.
- Once opened, it runs on your machine: the material never goes up to anyone's server.
1.What it changes and what it preserves
2 minWorth understanding the split before spending time installing, because it defines whether the tool fits your case.
- Change: background, lighting, clothing, and overall visual style of the scene.
- Preserves: facial identity, facial expressions, and character movement.
Spec sheet
- License
- Open source (code on GitHub)
- Resolution
- Up to 720p
- Model size
- ~80 GB
Continue in the full microcourse
You've read the opening of 3 classes
The microcourse covers the complete step-by-step, the selection criteria, where the tool fails, who it's really for — and, in the Expert version, the official address to start today.
- The path to getting it running2 min
- Where this fits into real production2 min
How we verified
We track releases straight from primary sources, transcribe what's demonstrated, check every name and number against the manufacturer's official documentation, and rewrite it in Portuguese — with what the tool no do it together, which is the part the ad leaves out.
