Which AI video model handles hands and objects best
Hands doing a task is where video models teleport objects and grow fingers. We asked 19 of them for one pour-over: a hand, a gooseneck kettle, a paper filter, a stream of water. The grid includes the September 18 additions; each model page dates its own run.
Anvisha Pai, Co-founder & CEO, Voyager
What the side-by-side shows
Kling 3.0 Pro produced the best pour: a steady hand, a kettle that stays a kettle and water that lands in the filter. Seedance 2.0, Wan 3.0 and Veo 3.1 are close. The rest reinterpreted the props: Veo 3.1 Fast and Hailuo 2.3 swapped the filter for a cup, MiniMax H3 turned it into a hand tilting a cup beside an espresso, Grok poured from above into a bowl held in one hand, and Wan 2.7 showed a whole café before finding the pour.
- 1.Kling 3.0 Pro a steady hand, a real kettle and water that lands in the filter.
- 2.Seedance 2.0 the pour on a glass carafe with steam, props intact.
- 3.Wan 3.0 a believable pour with the filter and kettle as asked.
- 4.Veo 3.1 steam, a gooseneck and a white dripper, held steady.
Every video model on this prompt, 19 clips
One try each, nothing re-rolled or retouched, generated by us. Ranked models come first; the rest follow in family order. Open a model for the rest of its set.
›The prompt every model got
Close-up of hands pouring hot water from a gooseneck kettle in slow circles into a pour-over coffee dripper, steam rising, shallow depth of field, warm cafe light.
Hands, a kettle and a steady pour without extra fingers or teleporting objects.
a steady hand, a real kettle and water that lands in the filter.
the pour on a glass carafe with steam, props intact.
a believable pour with the filter and kettle as asked.
steam, a gooseneck and a white dripper, held steady.
The other prompts
- Which AI video model holds an art style
- Which AI video model handles a busy scene best
- Which AI video model handles a camera move best
- Which AI image-to-video model keeps the product intact
- Which AI video model gets physics right
- Which AI video model makes the best product turntable
- Which AI video model makes the best talking head
- Which AI video model can render text on screen
Frequently asked questions
Can AI video do hands yet?
For a slow pour, yes: Kling 3.0 Pro, Seedance 2.0, Wan 3.0 and Veo 3.1 all kept the hand and the objects correct. The common failure is not extra fingers but swapped props, a cup for a filter.
Voyager
Run any of these inside Voyager
Voyager is an agent for creative work that runs every model on this page. Brief it, and it picks the model for the job, makes the thing and hands back something you can edit. Create an account to get started.