RENDASTIC GUIDES
How to Animate a Photo With AI: Step-by-Step Guide
Choose a clear photo, write a motion prompt, set the clip length and review your video. A practical guide with portrait, pet and product examples.
To animate a photo with AI, upload one clear image, describe what should move, choose supported video settings, and generate a short shot. The image supplies the composition; the prompt supplies the action. A small, readable movement is a more useful starting point than asking a single frame to become an entire story.
In Rendastic, start with image to video. The available free video profile is Rendastic 1.0 at 480p with a 4-second setting. Free use does not require an account, but it depends on your daily allowance, regional eligibility, current capacity and input checks. Other supported models and settings use credits.

1. Choose a frame that gives the movement room
Use a sharp JPG, PNG or WebP within the uploader's 10 MB limit. A single clear subject and an uncluttered background make it easier to describe the action. If you want a hand gesture, include the whole hand and enough arm. If you want a person to turn, avoid a crop that cuts tightly through the head or shoulders. Do not expect the model to reconstruct hidden details accurately.
A damaged scan, heavy blur or a tiny face gives the model more to invent. Repair obvious damage before animation, or choose another source. For products, leave a margin around the item and keep purchase-critical labels available in a separate still image. For two people or pets in one photo, inspect both subjects after generation rather than judging the first frame alone.
Upload only material you own or are authorized to use. Source images are processed by generation providers; this is not a local-only image editor. Read the upload and processing policy before choosing sensitive material.
2. Describe what changes, and what stays still
Begin with one subject action and, optionally, one camera movement. You usually do not need to rewrite the whole appearance of a subject that the image already shows. Instead, name the visible subject, state a small motion, and identify an important element that should stay still.
The person gives a small smile and gently turns toward the camera. Keep the camera still and the background unchanged.
This is an instruction, not a guarantee of identity preservation. Avoid packing walking, turning, waving and a scene transition into one short request. If you need a sequence, plan separate shots and assemble them in your video editor. If you need a specific dance to follow an existing performance, use a supported motion-reference workflow rather than expecting a short text instruction to reproduce choreography.
Camera language should also be consistent. A slow push-in means the camera moves closer. It is different from a subject walking toward the camera. Choose one for the first version so you can tell what needs adjustment afterward.
3. Check the controls below the prompt
The duration control sets the requested length. Writing “15 seconds” in the prompt does not override a 4-second selection. Each model supports its own duration, resolution and reference modes, so select a model with the actual controls your shot needs.
For the eligible free route, Rendastic 1.0 is presented as 480p and 4 seconds. A longer original video returned by that route is delivered in full; it is not cropped down merely to match the label. Generated audio is not guaranteed. For longer or higher-resolution work, review the supported paid settings and the credit amount beside Generate before confirming.
Framing can depend on the input mode. A starting-frame workflow may follow the source image's aspect ratio rather than an independent ratio control. Crop a copy of the photo to the composition you want before uploading if that mode locks the frame shape. Leave the original file untouched for later edits.
4. Submit once, then inspect the whole clip
After submitting, follow the task already shown in the editor or My Work. A slow response is not a reason to submit the same request again. If the task is awaiting confirmation, keep it attached to its original record; contact support with the task identifier if it remains unresolved.
When the result is ready, check the first frame, the movement and the final frame. Look at faces, hands, product edges, text and background geometry. A successful delivery means a video is available; it does not guarantee that every detail is suitable for publishing.
For another version, change one thing at a time: less movement, a steadier camera, a clearer image or a different supported model. Each new submission is a separate creation. Download the completed original within its seven-day availability window and keep your own copy. Rendastic adds no watermark to the delivered video; watermark-free delivery does not remove marks already in source material or replace rights clearance.
Three motion prompts to adapt
A portrait greeting
The person smiles softly and makes one small head turn. Keep the shoulders and background steady, with a fixed camera.
Choose a well-lit face with visible shoulders. Check facial details through the turn; use a smaller motion if the result changes the person's appearance too much.
A pet moment
The cat turns its head slightly and twitches one ear while remaining seated. Keep the camera still.
Use a source in which the ears and head are fully visible. Avoid asking for a leap from a tightly cropped seated portrait.
A product reveal
Slowly move the camera closer to the bottle while a soft light passes across the background. Keep the bottle in the same place.
Use a clean product photo with space around the edges. Add exact logos, prices and promotional text afterward in your editor if accuracy is essential.