Feedback
h3 recast
H3 Recast swaps the people in a video you have already edited — your portraits step in while the acting, cuts and audio stay as they were.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.

Seedance 2.0
The Future of AI Video Is Here.

Veo3.1
Create Stunning Videos with Veo3.1
FLUX 3 Video Generator

Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
H3 Recast: Swap the Performers, Keep the Performance
The tool works on material that is already in the can — it leaves the acting, blocking and sound mix alone and only changes who appears on screen.
- Casting Work, Not RegenerationNo part of the take is rebuilt from scratch — H3 Recast protects everything the footage already gets right and touches only the performer's identity.
- Up to Four New Faces, Ordered Left to RightHand over as many as four portraits and each one claims a leading figure in every shot that figure appears in, so a single identity carries across the whole edit.
- Acting, Cutting and Audio All SurviveEach glance and gesture, the rhythm of the performance, the way light falls on set and the final mix come through the swap without a scratch.
Running H3 Recast in Three Steps
Three short moves carry an edited clip all the way to a fresh cast with H3 Recast.
What H3 Recast Handles Behind the Scenes
Which parts of a clip H3 Recast carries over, which it replaces, and how to hold on to the original styling when that matters.
Identity Holds Across Every Cut
One portrait keeps claiming the same person in each shot that person turns up in, so the casting stays consistent as the edit jumps between cuts.
Prompt Control Over Cast Order
Pairing defaults to a left-to-right run; naming each on-screen figure alongside its photo in a short prompt overrides that assignment.
Face-Only Swap on Request
With an empty prompt the portrait's styling may travel along with the face; ask for the face alone when the original hair and wardrobe must stay untouched.
Low-Res Check Before the Final
A fast preview render settles whether the new cast reads correctly before any credits go toward the full-resolution finish.
Runtime Mirrors the Source
What goes in is what comes out in length, so spend tracks how long the clip actually runs rather than a setting you have to guess at.
Pacing and Sound Come Through
The rhythm of the edit and its audio arrive untouched, which means the new version drops straight back into the timeline.
Questions About H3 Recast, Answered
Straight answers on swapping the people in a clip that is already finished, and on how H3 Recast behaves along the way.
How long can the source clip run?
Anywhere from five to thirty seconds, and no single shot beyond fifteen. Cut longer takes apart first; the output matches the input in length.
How many people can be swapped, and in which order?
Four at most, one portrait each. Leading figures are claimed left to right by default, and a prompt can reassign that pairing.
Which reference photos give the best result?
One front-facing portrait per person, sharp and evenly lit. Avoid sunglasses, hats, covered faces, group shots, soft focus and heavy filters.
Will hair and clothes change as well?
Frequently yes when no prompt is given, even from a head crop. Request the face only if the original hair and wardrobe have to survive.
Preview at 768p or go straight to 1080p?
Check the casting at 768p first, then finish at 1080p; what you pay follows clip length.
When is this the wrong tool?
When new movement, camera work or a new location is required. Crowds, small faces, fast action and occlusion break it, and likeness can drift across long shots.
Put H3 Recast to Work on Fuser
Bring a clip that is already cut and scored, then change only who appears inside it — timing, camera work, edits and audio come out the other side untouched.
