Start with a clip
Choose a continuous, 3–10 second MP4. One visible character and a clear performance work best.
INDEPENDENT STUDIO / INVITATION BETA
BETA ACCESSKnow the cost.
Then make the cut.
Every generation has a clear credit quote. During beta, credits are granted by invitation. Public credit purchases are not open yet.
VIDEO + IMAGE → NEW VIDEOPOWERED BY KLING
THE TRANSFORMATION ROOM● REC
TAKE
02
02
ACTUAL STUDIO OUTPUTMake it
your own.
GENERATED TAKE / KLINGyour own.
↳ SAME IDEA. NEW DIRECTION.16:9
A DIFFERENT REALITY ↗
INVITATION BETATry the studio.
Try the studio.
Find your next idea.
Sign in with an email link, then request beta access. We grant credits manually while we test quality and reliability.
Request beta credits ↗CLEAR QUOTES
63–416 credits / video
Economy motion transfer: 105 credits for a 5-second clip. Kling 3.0 motion transfer or Kling O1 character swap: 189 credits for 5 seconds. Credits are calculated at 300 per $1 of estimated provider cost.
We bill in whole seconds, rounded up. The exact quote appears after your video upload. Credits are held when a job starts and refunded if it fails.
Public credit packs and subscriptions are not available during this beta.
THREE STEPS. ONE NEW TAKE.You bring the footage.
You bring the footage.
We bring the possibilities.
Give it a new identity
Add a character image. Choose character swap to edit the shot, or motion transfer to animate your reference.
Make your next version
Review your credit quote, generate, and download the result. Refine your references to improve the next take.
BEFORE YOUR FIRST TAKEA few things
A few things
to know.
What is Genjutsu Video?
An independent video editing tool using third-party AI models. Character swap runs on Kling O1, and motion transfer uses Kling 2.6 or Kling 3.0 Motion Control via fal.ai.
Is this Higgsfield’s Genjutsu?
No. Genjutsu Video is a separate product and is not affiliated with Higgsfield. Our workflows and outputs differ from Higgsfield Genjutsu.
What kind of files can I upload?
Standard MP4 videos between 3 and 10 seconds, up to 30 MB, with both dimensions between 720 and 2160 pixels. Use 24–60 fps. Character references can be JPG, PNG, or WebP up to 10 MB.
Will the movement stay exactly the same?
The models use your video to guide movement. Results can change details, timing, or background elements. Fast actions, occlusions, and multiple characters may reduce consistency.
Can I try it for free?
We are currently in invitation beta. Sign in and request beta credits. No subscription or payment is required to request access.
What happens if a generation fails?
Credits held for a failed job are returned automatically. A completed video with an unexpected aesthetic result is still a completed generation.