Image to video AI takes one photo and turns it into a few seconds of video: waves start rolling, a person turns toward the camera, steam curls off a cup. You upload the photo as the first frame and describe what should happen next. In Toybox AI, that job is the "Start from a photo" option in Video Creator, and the lower-cost Video Creator Lite has the same option. The sections below match a motion to each kind of photo, give six prompts to copy, and cover the two free Google routes, Toybox's price per clip, and where faces and lettering still slip.
What you can make with image to video AI
Any photo with one clear subject can move. These kinds of photo suit it, each with a motion that fits:
- Product photos. Steam rising from a mug, light sliding across a glass bottle, a sneaker turning slowly on a stand. If all you have is a phone snapshot, stage the product first; the guide to product mockup videos walks through both steps.
- Old family photos. A slow blink, a small smile, a head turning toward the camera. The page on how to animate old photos covers scanning, colorizing and keeping the motion gentle.
- Travel and landscape photos. Waves washing up the sand, clouds drifting, lanterns swaying. Water, sky and leaves are forgiving to animate, because a wave that comes out a little different still looks like a wave.
- Flyers and posters. Confetti falling past the date, stage lights sweeping behind a band name. The lettering is the risky part, and animated flyers covers motion that keeps the words readable.
- Pet photos. A dog's ears perking up, a cat's tail flicking, a head tilt toward the camera.
- Drawings and paintings. A crayon rocket lifting off the page, rain starting to fall in a watercolor. Google's Veo documentation lists animating drawings and paintings as one use of image-to-video.
- A move from one photo to another. With an end frame, the clip has to finish on a second image you choose: the same street at dusk and after dark, an empty cake stand and a finished cake.
How image to video AI works
An AI video generator from image input, such as Google's Veo, treats your photo as frame one and generates the frames that follow, guided by your prompt. Google's Gemini API documentation says Veo starts the video on the image you supply, and that Veo 3.1 makes 4, 6 or 8 second clips at 24 frames per second.
Everything the camera never captured has to be invented. When a person turns, the AI draws the side of a face it has never seen. When the camera pulls back, it draws the rest of the room. That's why small motions look more like your photo than big ones.
An end frame gives the model a second fixed point. The clip starts on your first image, ends on your second, and the prompt describes what happens in between. In Toybox, both video tools take an end frame once a start frame is added, and an end frame makes the clip 8 seconds long.
Motion prompts that make a photo move
To make a photo move with AI, name what moves, what the camera does, and what has to stay put. Paste each prompt below into "Describe your video" in Video Creator after adding your photo as the start frame, then swap in the details of your own picture.
Portrait with a small smile
Use this for a head-and-shoulders photo of one person who agreed to be animated.
The woman in the photo blinks slowly, then a small smile spreads across her face as she lifts her chin slightly toward the camera. A light breeze moves a few strands of her hair. Static camera at eye level. Her face, hairstyle, clothing and the background stay unchanged from the photo. Soft natural light.Product on a table
Use this for a product shot with some empty space around the item.
Steam rises gently from the coffee mug on the wooden table, curling and fading into the air. Morning sunlight slides slowly across the table from left to right. Very slow push in on the mug. The mug, its color and its printed label stay exactly as photographed. Calm, warm cafe mood.Beach or travel photo
Use this for a wide shot with water, sky or trees in it.
Waves roll onto the beach and slide back, leaving a thin sheet of water on the sand. Clouds drift slowly to the right, and the palm leaves sway in a light wind. Slow pan to the right. Keep the horizon level and the colors of the sunset the same as the photo.Pet photo
Use this for a pet looking at, or near, the camera.
The dog in the photo tilts its head to one side, its ears perk up, and its tail wags twice. Keep the dog's markings, its coat color and the living room behind it the same. Static camera at the dog's eye level, shallow depth of field. Playful, cozy mood.Child's drawing
Use this for a scanned or photographed drawing, shot flat and straight on.
The crayon rocket in this child's drawing lifts off the paper launchpad, trailing scribbled orange flames, while the drawn stars twinkle. Keep the crayon texture, the white paper and the hand-drawn style. Static camera looking straight down at the paper.Two photos, day to night
Use this with a start frame and an end frame taken from the same spot.
Time passes on this street corner from late afternoon to night: the light turns golden, then deep blue, and the shop signs and street lamps switch on one by one. Locked-off camera, no movement. Keep the buildings, the street and the parked bicycle in the same place from start to end.Tips for better image to video results
- Match the photo's shape to the video. The two formats are Wide (16:9) and Vertical (9:16), and Toybox won't crop or resize your start frame. Crop the photo before uploading: 9:16 for Shorts, Reels and TikTok, 16:9 for YouTube and websites.
- Give it one motion and one camera move. "The flag ripples, slow push in" is a complete brief. If a clip comes back chaotic, cut motions from the prompt until one is left.
- Name what must not change. End the prompt with "Keep her face, clothing and the background exactly as in the photo." Toybox doesn't guarantee that faces, text or labels stay unchanged while they move, so watch for it.
- Use real camera words. Google's Veo 3.1 prompting guide names camera moves including the crane shot, the POV shot and the aerial view. Write "static camera" when only the subject should move.
- Start from the clearest original you have. A screenshot of a photo or a photo of a screen gives the AI blur to animate. For a blurry original, try Image Unblur first, then look closely at every face next to the original, because the sharper detail is the AI's reconstruction, not what the camera caught.
- Pick the length for the motion. A blink or a smile fits in 4 seconds. A slow pan across a landscape needs 8.
- Name what you don't want. Open More options in Video Creator and fill in "Things to avoid", for example "text on screen, extra people, fast camera". Video Creator Lite doesn't have this field.
What image to video AI can't do
- Make a long video in one go. Each clip is 4, 6 or 8 seconds. A longer video is several clips joined in an editor; the guide on how to make AI videos covers building a sequence.
- Keep every face, hand and letter exact. Features can drift and small text can warp as the image moves. Keep the motion small, watch the clip at full screen, and create it again with less motion if something changes.
- Give you sound controls. Video Creator and Video Creator Lite clips come with sound, but there's no volume or music setting. Describe the sounds you want in the prompt, or add your own voiceover or music track in an editor; the guide to AI video with sound covers both.
- Animate people without their say. Toybox's Terms of Service prohibit generating images of real people without their explicit consent.
- Run on the Free plan. Both video tools need a Pro plan or higher. A Video Creator clip currently costs 350 credits at any length or shape. Video Creator Lite currently costs 180 credits at 720p and 220 at 1080p, and 1080p clips are always 8 seconds. Plans currently start at $9.99 a month for 1,000 credits, which covers two Video Creator clips (700 credits) or five 720p clips in Lite (900 credits). See current prices.
Free ways to make a photo move
Two Google apps let you animate a photo with AI free of charge, up to a daily limit. Both suit a quick test before you pay for anything.
- Google Photos, Photo to video. In the Google Photos app, open Create, pick Photo to video, and add the picture. "Subtle movements" adds small motions such as a wave or a smile, and "I'm feeling lucky" can add bigger ones, like dancing or confetti. Typing your own prompt requires being 18 or older. Each result is a six-second portrait clip, so a landscape photo loses its sides. The number per day is capped (a Google AI plan raises the cap), the photo must be backed up, and a personal Google Account is required (work and school accounts are excluded).
- Google Flow. Flow gives 50 credits a day to users 18 and older in the countries where it's offered, with or without a Google AI plan. Set the prompt box to Video and then Frames, put your photo in the start frame slot, and type the motion. At 10 credits per Veo 3.1 Lite clip, that's five tries a day. Without a Google AI plan, Flow may refuse video requests at peak times, between about 2 PM and 5 PM UTC.
The YouTube app also has a Photo to video option for Shorts, outside the United Kingdom, Switzerland and the European Economic Area. For paid image-to-video tools beyond Google's own apps, see Runway alternatives.
Video Creator or Video Creator Lite for your photo
Video Creator is powered by Google Veo 3.1, and Video Creator Lite uses Veo 3.1 Lite. Both take a start frame and an end frame, and both make Wide or Vertical clips. Pick by the job, with prices as of September 2026:
| Your photo and goal | Tool and settings | Credits per clip (currently) |
|---|---|---|
| Testing a motion idea before a final version | Video Creator Lite, 720p, 4 sec | 180 |
| A vertical clip for Reels or Shorts at a higher resolution | Video Creator Lite, 1080p, Vertical (8 sec) | 220 |
| A portrait or product where you want to list things to avoid | Video Creator, "Things to avoid" filled in | 350 |
| A move from one photo to a second photo | Either tool, start and end frames (8 sec) | 180 to 350 |
Video Creator makes 720p clips. Video Creator Lite offers 720p or 1080p. To animate a photo into video at the lowest cost, open Video Creator Lite at 720p.
How to make a photo move in Video Creator
- Prepare the photo. Crop it to 16:9 or 9:16 and save it as a JPG or PNG under 10 MB.
- Open Video Creator and sign in. It needs a Pro plan or higher.
- Select "Start from a photo" and pick your image. It becomes the Start frame. To fix the final shot, add a second image as the End frame.
- Describe the motion in "Describe your video". Say what moves, what the camera does and what stays the same, in up to 2,000 characters. The mic (Chrome, Edge or Safari) lets you speak it instead.
- Choose the Format and Length. Wide (16:9) or Vertical (9:16), and 4, 6 or 8 sec. With an end frame, the clip is 8 seconds.
- Select "Create video". Generation takes a few minutes. If a generation fails, your credits go back automatically.
- Check it before you share it. Watch at full screen and hold faces, hands and any lettering up against your photo, then download the MP4, or keep it in your Toybox gallery.