THE IMAGE-TO-VIDEO GUIDE
How to turn a photo into a video
Turn an existing portrait, product photo or scene into a short clip without arranging a video shoot. This guide walks you through choosing an image, writing a motion prompt and downloading a video you can finish for a social post or website.
Choose your image ↗Generation requires sign-in and sufficient credits. One-time packs start at $9.90.
Choose between AI motion and a simple pan or zoom
If you only want to move across a photograph, use a pan-and-zoom effect in a video editor. Set the starting crop, ending crop and clip length, then export. This moves the existing image without asking a model to invent a blink, new background or changing object.
Choose AI image-to-video when you want new movement inside the scene, such as a person blinking or fabric moving. Review the result because the model can redraw details. Frovimo generates AI motion; it does not offer a separate deterministic pan-and-zoom export.
| Your goal | Start with | Check before publishing |
|---|---|---|
| Move across a still photo while keeping its contents | Pan/zoom in an editor; use the same image at both ends. | Crop, export resolution and compression. |
| A gentle blink in a portrait | One short AI request with a fixed head and camera. | Eyes, mouth and identity throughout the clip. |
| A product clip for a vertical post | One AI movement, followed by an editor for captions and a call to action. | Label, silhouette and readable text. |
| An unchanged product with smoke behind it | Separate product and effect layers in an editor. | Mask edges, occlusion and brand details. |
| A photo speaking a recorded script | A talking-avatar workflow with audio and lip-sync controls. | Consent, voice rights and mouth synchronization. |
1. Choose a clear starting image
Start with a subject you can see clearly and enough space around it for the movement you want. If the camera moves closer, parts near the edge may leave the frame. Choose your intended framing before spending credits.
Look closely at faces, hands and product lettering in the source. Small or blurry details are already hard to judge; a video should not be relied on to restore information missing from the image.
Upload a JPG, PNG or WebP file up to 8 MB. Use material you own or have permission to process. Generation sends the image and prompt to cloud services; read the privacy policy before uploading a sensitive photo.
2. Describe one movement
The image already supplies the scene. Use the prompt to describe what should change: a subject action or a camera movement. Then name what should remain still.
For a product photo, try: Slowly move the camera closer to the product. Keep the product still and keep the label visible. This asks for a camera move while keeping the product still. It guides generation; it does not lock the lettering or guarantee an exact result.
| Starting idea | More specific first attempt | What to inspect |
|---|---|---|
| “Bring this portrait to life.” | The person blinks gently. Keep the head, expression and camera still. | Eye movement, mouth and face shape. |
| “Make a dramatic product ad.” | Slowly move the camera closer to the product. Keep the product still and keep the label visible. | Product outline, lettering and crop. |
| “Animate everything in this landscape.” | Slowly pan the camera to the right across the scene. Keep the buildings and horizon stable. | Buildings, horizon and frame edges. |
Use these starting prompts with your own image. Change one instruction at a time to explore different movements and compare the results.
Choose a starting prompt on the homepage ↗3. Choose settings and a revision budget
The current Grok Imagine workflow offers 6–30 seconds at 480p or 720p. Ratios are 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. Check the studio's available settings and quote before submitting.
| Setting | How to choose a first attempt | Tradeoff to check |
|---|---|---|
| Duration | Use 6 seconds to inspect one simple movement. | A longer request costs more and gives you more footage to review. |
| Resolution | Use 480p for a lower-credit draft; choose 720p if that is the resolution you need. | A new 720p generation is a separate attempt, not an upscale of your draft. |
| Framing | Try 9:16 for a vertical post, 16:9 for a landscape clip or 1:1 for square. | Changing the ratio can change the composition. Check the subject and frame edges in the result. |
A 6-second request uses 10 credits at 480p or 20 credits at 720p. Each second beyond six adds 4 credits. Each new attempt spends credits, even if you use the same image and prompt.
Compare pack prices and request budgets before buying. A quoted portion of a credit pack is not an individual clip purchase.
How much control does a motion prompt give you?
A prompt describes the movement you want. It does not pin pixels, mark a motion path or lock a region of the image. The current Grok Imagine workflow accepts a starting image, prompt, duration, resolution and framing; it has no motion brush or reference-video control. Start/end frame availability depends on the active model and is confirmed in the studio.
For a portrait, start with The person blinks gently. Keep the head, expression and camera still. rather than asking for a blink, smile and head turn together. Use a photo with a clearly visible face and permission from the person pictured. Check the eyelids and mouth at normal speed, then pause anywhere the face changes.
If the face drifts, remove extra actions and keep the camera fixed for your next attempt. Keep the other settings unchanged so the comparison is useful. Each new request costs credits, and a simpler prompt still cannot guarantee identity preservation. If the next result is unsuitable, reconsider the workflow before spending on more retries.
Keep exact product details with separate layers
If the product, label or logo must remain unchanged, a whole-image generation is a poor fit for that requirement. A prompt asking for “unchanged text” is not a guarantee. Adding a logo afterward can repair the logo, but it cannot fix a distorted product underneath it.
- Keep the original product image on its own layer. Mask or remove the background in your editor and check fine edges.
- Place a smoke, light or background-motion clip on another layer. Generate that effect separately or use footage you have permission to use.
- Composite the layers. Keep the product layer still when its shape must remain fixed; put effects behind it where appropriate.
- Add exact lettering, captions and a call to action as editor layers. Review the full export for mask edges, unwanted overlap and readable text.
This is an editing workflow, not an automatic feature in Frovimo. It keeps the original product content in the edit; scaling, cropping and compression can still change exported pixels. For a looser product preview, try the short product setup and inspect the result before treating it as an advertisement.
4. Watch the entire clip, then make one change
Check the start, middle and end, then watch the whole video. Look at the details that matter for your use, rather than judging only the opening frame.
| If you see this | What to change for another attempt | What still needs checking |
|---|---|---|
| A portrait changes too much | Remove a head turn or smile. Request a blink with a still head and fixed camera. | Eyes, mouth, face shape and hands throughout the clip. |
| The product moves when you wanted a camera move | Ask the camera to move closer and explicitly keep the product still. | The product outline, labels and logos. Exact lettering may still need an overlay in your editor. |
| The scene changes or important details leave the frame | Reduce the camera movement. Name the buildings and horizon that should remain stable. | Architecture, horizon and crop from start to finish. |
These are suggestions for a new attempt, not guaranteed fixes. To compare a prompt change, keep the image, duration, resolution and ratio the same. Outputs can still vary between requests.
The saved balcony clip on the homepage gives you something to inspect: fingers meeting fabric, shirt shapes and railings. The creator provided this AI-generated example; the page includes its motion prompt and measured video dimensions.
5. Download and finish your clip
When the request is complete, preview and download the returned video from the studio or your history. Save a local copy of the files you want to keep.
Frovimo adds no site watermark to the returned file. Marks in your input or the provider's output are not removed. Add captions, exact brand overlays, music or a voiceover in a video editor. The current Grok Imagine workflow has no voice upload, audio toggle or lip-sync control.
Before publishing, confirm you have the necessary permissions for people, brands and other material in the image. Buying credits does not grant those rights or guarantee commercial suitability. Review the usage terms.
If a request does not finish
Check the same request in the studio or History while it is pending. Checking status does not start a new generation or spend another generation's credits. A new submission is a separate paid attempt.
A pending request is not a confirmed failure. When a request is recorded as failed, the service attempts to return its credits. Check the error and your balance. A credit return does not reverse a purchase.
For a missing result or credit return, contact support@ai-image-to-video.com with the request reference, its status and a description of the issue. For a purchase issue, include the order reference instead.
Try one movement with your image
Choose a starting photo, set your budget and inspect the result before making another request.
Upload an image ↗