Travel and landscapes
Choose a sharp scene with a clear foreground and background. Water, clouds and trees give you a specific motion to describe. Start with a slow push toward the horizon rather than a fast flight through the scene.
Photo to Video AI
Turn a travel photo, product shot or portrait into a short AI video. Upload your photo and describe the movement you want.
Which photos are worth trying?
Start with a still JPEG, PNG or WebP under 20 MB that you have permission to use. Pick a crop close to the shape of your intended video: wide for 16:9 or tall for 9:16. Leave space around the subject so a camera move has room to work. Blur, heavy compression and crowded scenes make small details harder to preserve.
Product photos
Use one clearly lit product against a simple background, with space around its edges. Try a small camera move or changing reflections. Tiny labels, logos and exact packaging shapes need close review in the output.
Portraits and memories
Choose a clear face with even lighting and an uncluttered background. Begin with a gentle camera move rather than a large head turn. AI can change identity and expression; an animated memory is an interpretation of the photo.
Describe the motion, not just the mood
Your photo supplies the starting scene. Use the prompt to explain what moves, in which direction and how the camera behaves. Start with one action over a short clip. Add sound instructions only when they matter to the scene; generated audio can also vary.
A quiet travel shot
The camera slowly pushes toward the distant shoreline. Small ripples move across the water. Keep the horizon level and the movement gentle.
Use a photo that already contains water and a visible horizon. Asking for a new location gives the model more to invent.
A product detail
The camera makes a small move from left to right around the bottle. Soft studio light moves across the glass. Keep the bottle upright and the background simple.
A small move reveals less unseen surface than a full orbit. It still cannot guarantee an unchanged logo or label.
A gentle portrait
The camera slowly moves closer to the person. A light breeze moves a few strands of hair. Keep the head still and the expression relaxed.
Treat this as a prompt to test. Check the face throughout the clip before sharing; this page has no portrait demonstration.
These are starting prompts to adapt to your own photo. Instructions guide the model; they do not guarantee an unchanged subject.
Two real generations and what they reveal
These are actual H3 Max Turbo outputs made from our own illustrations, rather than photographs. Both use one image, a 5-second duration and 480p settings. They show motion and detail changes to look for when testing your photos; they do not establish how a portrait or your own image will perform.
Landscape motion: allow room for change
View the exact generation prompt
The camera drifts forward over the water as the sun rises. A soft breeze, gentle ripples, distant birds.
The output adds motion to the lake scene, but it also changes the size of the sun and the arrangement of the landscape. A motion prompt does not lock the original composition.
For a travel photo, start with a modest move and check the skyline, buildings and landmarks. Use the original photo when exact documentary accuracy matters.
Product motion: check the lettering
View the exact generation prompt
A slow orbit around the bottle. Light glides across the glass, with subtle reflections and a clean studio background.
The generated bottle has a more photographic appearance than the source illustration. Its proportions and label details change, so this clip is a creative concept rather than proof of packaging accuracy.
For a real product photo, inspect the logo and label frame by frame. Reduce the camera move if the product changes; use filmed footage for a shot that must preserve every detail.
Make your first photo video
Upload one clear photo
Select your image in the generator above. Start with one photo so you can learn how the model interprets that scene before adding more references.
Choose one movement and check the cost
Adapt a prompt above, select the video shape and review your settings. H3 Max Turbo at 5 seconds and 480p costs 10 credits for one image. Longer clips, other models and multiple references cost more; the generator shows the exact quote.
Generate, review and download
Sign in with Google when prompted, then confirm the generation. Review the full video for unwanted changes before downloading it. Keep the original photo so you can compare the subject and decide whether another attempt is worth the credits.
Common problems and useful next steps
Faces, hands or labels change
Try a clearer crop, less subject movement and a simpler background. Review the whole video, not only the first frame. Prompt instructions cannot guarantee identity, lettering or object geometry.
The scene bends or invents new objects
Remove conflicting actions and ask for one small camera move. A large orbit reveals areas that the photo never captured; the model has to imagine them. A more detailed prompt will not recover missing evidence.
The motion is weak or goes the wrong way
Name the moving subject, its direction and the camera behavior. Replace “make it cinematic” with “small waves move toward the shore; the camera stays still.” Adjust one instruction at a time so you can compare results.
Upload or generation does not complete
Use a still JPEG, PNG or WebP under 20 MB. Convert HEIC or animated files before uploading. A confirmed generation failure returns reserved credits; a delayed provider response keeps the same task under review. Check that task before submitting another paid generation.
Before you animate a photo
Is this a photo slideshow maker?
This workflow generates new motion within an image. It does not arrange an album into timed slides or add an uploaded music track. Multiple images act as visual references for one generated video, rather than a sequence of slideshow cards.
Will AI preserve every detail in my photo?
No. Faces, text, geometry and background details can change. Use gentle motion and review the output. For evidence, faithful product documentation or a memory that must remain exactly as photographed, keep the original image or use filmed footage.
Can I animate several photos together?
The generator accepts up to nine images as references for one video. You can also select a first and last frame. Reference images and selected frames cannot be mixed in the same generation. Begin with one photo when testing a new scene or prompt.
Are my photos published as examples?
Your library is private by default and your uploads are not automatically published as examples. Infrastructure and model providers process the images needed for generation. Read the privacy policy for storage and deletion details.
Choose a photo and try one small movement
Start with a scene you know well, so you can spot what changes. Prepare your photo and prompt, then use the generator above.
Create a photo video