Kling O1 Released, Claimed as a Unified Multimodal Video Model, Editing Video Like Editing Images

资讯 / AI 快讯
1

Kling just dropped O1, claiming it’s the world’s first unified multimodal video model. The line that grabbed me the most was, “Make editing video as easy as editing an image.” For storyboard artists, the biggest headache is revisions — a client says a shot needs to move a little more, and in the traditional workflow, you basically start from scratch. If it can really do local video editing like adjusting layers, that’s a huge shakeup for our industry.

I’m holding off on the whole “unified” claim — video adds the dimension of time compared to images, so keeping frames consistent and not breaking the whole clip when you tweak one spot is the real challenge. The name O1 is clearly aiming for flagship status. I’m gonna throw in a few shots that need constant revisions and see if it’s just marketing hype or if it can actually deliver.

2

P videos like P images — if that were real, it’d be a total game-changer.

1 Like
3

Sitting here waiting for the actual test results.

4

The real benchmark is fixing one thing without breaking the whole piece—agree with the storyboard take.

5

That whole “one model to rule them all” thing sounds nice in theory, but let’s be real—once you add video into the mix, the time dimension is a total nightmare to handle.

6

changing one spot without breaking the whole thing, I’m fully with that, right now moving one shot means regenerating the whole segment which is torture