Who is in the frame
One person or two, how close they stand, who faces the camera. Faces are the part viewers check first, so decide this before anything else and crop the source photo to match.
There is usually a picture like this and rarely a video. Send in the one where she is looking down at you — a phone shot, a scanned print, a frame from the hallway — and name the single thing that moves: she lifts her head, the wind takes her hair, she smiles at whoever is holding the camera.
Animate a Photo









Prompt Guide
Model
Reference Image
Three steps, one gesture.
Drop in a phone photo, a scan or a screenshot. If her face is smaller than a thumbnail in the frame, crop closer first — soft faces are almost always a framing problem, not a model problem.
Say what changes in the clip: she lifts her head, he steps in for a hug, the candle gutters. One action and one camera move read as real; three actions read as a slideshow.
Choose the frame you need — vertical for a phone, wide for a slideshow — and reroll if lips or hands drift. Keep the version you would actually send to family.
Different moods, different prompts. The words beside each clip are the ones that produced it.
A woman in her sixties looks up from breakfast and finds a small bunch of flowers beside her plate; her face softens into a smile. Bright kitchen, soft daylight, stable close framing.
Try This StyleA man in his thirties hands his mother a wrapped bouquet on a sunny porch; she takes it with both hands and laughs. Waist-up composition, warm morning light, gentle push-in.
Try This StyleAn old framed photograph of a young mother and her small son comes quietly alive: she lifts her head a little, he blinks. Warm film grain kept, original framing held, motion barely there.
Try This StyleA paper-cut mother and son wave from a handmade garden gate while layered paper leaves stir in the breeze. Warm pastel palette, soft shadows, gentle stop-motion feel.
Try This StyleEverything else is decoration.
One person or two, how close they stand, who faces the camera. Faces are the part viewers check first, so decide this before anything else and crop the source photo to match.
A five-year-old and a mother in her thirties read differently from a grown man and a woman in her sixties. Naming the age changes the faces more than any style word does.
Window light, candle, late sun, overcast beach. The light decides whether a clip looks like a photograph that moved or a photograph that was redrawn.
A head turning, a hand arriving, a flame guttering. Pick one; two competing gestures is the usual reason a clip looks artificial.
Vertical for a status, square for a printed card, wide for the slideshow at a family dinner. The same idea usually survives all three.
The clip downloads to you. The tool does not post, share or publish anything on your behalf.
I sent my mum a three-second clip made from a photo I took at her birthday in 2019. She watched it on repeat and then asked me how I did it. It took two rerolls before the smile looked like hers and not like a stock smile.
The free run is what got me to try it at all. I put in a photo of my grandmother, saw what the default model did, then switched to a second model for a cleaner result. Being able to compare before paying is the part I tell people about.
We lost my mother in 2021 and had almost no footage of her moving. I made one quiet clip from a portrait and kept it on my phone rather than posting it. For our family that was the right amount of this.
Small movements keep the face hers: a head turning, a blink, a smile arriving. Big gestures and long clips give the model room to improvise, and hands and teeth are where it shows first. If the result reads as somebody else, reroll — a second take is usually closer, and nobody has to see the first.
Plenty of families do it, and most treat the clip as a keepsake rather than a recording. Two practical things help: use photos you have the right to use, and keep the result to yourself until the people closest to her have seen it. What comes back is an imagined moment, not footage of something that happened — worth saying out loud before it gets shared around.
That is the normal starting point, not a bad one. Grain and creases tend to survive into the clip, which most families prefer to a face that looks newly invented. The real limit is size: if her head fills only a corner of the frame, crop in first, because soft faces are almost always a framing problem rather than a model problem.
One photo of the two of you together is the reliable route. With two separate portraits you can describe the shared frame — who stands where, how close, what the room looks like — but the model has to invent the join, so expect to reroll a few times and expect close family to notice. Most people outside the family will not.
You get one free run on the unlocked model, Grok Imagine Video 1.5. Every other video model in the picker is paid and marked locked, so you can see what you are choosing before you spend anything. Start with the free run to check that your photo and your one movement behave, then move up only if you want a different look.