Put two people in the same frame
They never have to have been photographed together for the clip to work.
AI Hug Video Generator
Turn 2 photos into heartwarming hugging videos. Choose Warm, Reunion, or Lift styles with realistic, touching motion.
First person
*Photos of consenting adults only
Second person
*Photos of consenting adults only
Style
Upload 2 photos to start
Before & after examples
Before
After
Before
After
Before
After
Before
After
Before
AfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterBeforeAfterWhat people use the AI hug generator for
Two photos, one scene. No prompt writing.
Make a card that moves
A short hug clip does more than a still for a birthday, an anniversary or a message.
Reunite people who are apart
Two separate photos become one moment of them meeting.
Keep both faces recognisable
The references are held across every frame, so it stays the two people you uploaded.
Start with a still, then animate
Use image output to lock framing and faces, then switch to video when you want the hug to move.
Set the mood in one click
Warm, reunion and film each direct the pacing and the light differently.
How it works
- 1
Upload two photos
Upload a photo of each person.
- 2
Continue in Studio
Choose image or video, pick a style, check the displayed cost, then select Create hug. Continue in Studio and sign in if asked; your photos and draft are carried with you.
- 3
Check and download
Watch and download your clip when it finishes.
Why the result holds up
Both faces are held across the clip
Identity drift is what breaks two-person video: by the third second one face has become someone else. The prompt pins both references for the whole clip, which is the hardest part and the reason this is not just a text prompt.
The motion is directed, not left to chance
Walking toward each other, the pause, the embrace — the beats are described so the clip has a shape instead of being five seconds of drift.
Framing and light are chosen for you
Camera movement, pacing and lighting come from a prompt we maintain. You choose a mood chip if you want one; there is nothing else to set.
This tool or the full studio?
Both run on the same models and the same credits. The difference is how much of the decision you want to make yourself.
AI Hug GeneratorImage / Video Studio
You supplyTwo photos · plus an optional line of your own to steer itYour own prompt, plus negative prompt, seed and batch size
You chooseThe built-in prompt does the rest — you can layer your own note on topAny model, quality tier, aspect ratio, duration and resolution
Best whenYou want two people in one moment without directing the motionYou know what you want and need control over how it is made
Why run it on Lumi Arts
Yours by default
There is no public gallery, no feed, and no discovery page. Uploads and results are visible only to you, and you can delete either at any time.
You see the price before you spend
Every run shows its exact credit cost on the button. No subscription is required — credits can be bought as a one-off pack.
Nothing to configure
The prompt, the model and the settings are chosen and maintained by us. You upload, pick a style if you want one, and go.
It keeps working after the tool
Results land in the same workspace as everything else you make, with your history, downloads and the full image and video studios one click away.
Frequently asked questions
How does the AI hug generator work?
You upload two photos, one per person. The model animates them moving toward each other and sharing a warm, natural hug, holding both faces consistent with your references across the whole clip. Camera movement, pacing and lighting are directed by a prompt we maintain.
How much does one run cost?
Image runs use the image model tier shown on the button. Video is five seconds at 768p and is billed per second by the model provider, so it costs more than an image. The exact price is always shown before you start.
How long does it take?
Usually one to three minutes. You can leave the page while it renders; the result appears in your workspace and in your task history when it's done.
What photos work best?
Clear, well-lit photos where each person's face is unobstructed and roughly front-facing. Full-body or upper-body framing gives the model more to work with than a tight head crop.
Whose photos can I use?
Your own, or photos of adults who have agreed to it. Images of minors are prohibited and blocked.
Can I get a still image instead of a video?
Yes. Switch the output to image on this page or in Studio. Video remains available when you want a short clip.
Why does video cost so much more than an image?
The model provider bills video by the second, and a five-second clip is many times the work of a single frame. We pass that through rather than averaging it into the image price — the exact cost is always on the button before you start.
Same-theme guides and workflows
More AI Studio tools
Two photos, one hug
Prepare one clear photo per consenting adult. Choose image or video and a style here, then continue in Studio to sign in if needed and create your result.
Create hug