Options below come from the same model catalog as Create. Setting combinations and credit costs are confirmed there before generation.
The model list is not available here yet. Open Create to see available models and settings ↗
How to compare your first two takes
Use the same source image and motion prompt with each model. Compare subject detail, how closely the movement follows your instruction, and the overall framing. A model that works well for a portrait may give a different result on a busy product scene.
Check duration and resolution before each attempt. A longer clip or a larger output is not automatically a better match for a simple gesture, and the credit cost can change with your settings.
Image to video questions, answered.
What is AI Image to Video?
It generates a video from a starting image and a description of the movement you want. Unlike Text to Video, you provide a visual reference for the scene.
Where do I upload my image?
Use the upload or animate button to open Image to Video in Create. Upload your picture there, choose a model, and review the settings before generating. Any motion prompt you write on this page carries over.
Which pictures work best?
Start with a sharp image and a clear subject. Avoid heavy blur or very small source images. For portraits, use a well-lit face; for products, make sure the important details are visible.
What should I write in the prompt?
Describe what should move, how it should move, and what the camera should do. Try one simple action first. You can ask for a steady camera if you want the subject movement to be the focus.
Will the image stay exactly the same?
Not necessarily. AI generation can change faces, text, shapes, and other details. Review the result before using it, especially for product images or portraits where accuracy matters.
Which durations, resolutions, and audio options are available?
Available settings depend on the selected model. Create shows the options for that model before you submit a generation.
How much does Image to Video cost?
The credit cost depends on the model and settings. Review the displayed cost in Create before generating, or visit Plans to compare current credit allowances.
How long does generation take?
Timing varies with the model, video settings, and queue. You can check the request status in Create while your video is processing.
Can I use the same image with different models?
Yes. Keep the source and motion prompt the same while comparing models. Each model can have different duration, resolution, audio, and input settings, so check them before submitting another generation.
Can I use a first and last frame?
Image input support varies by model. Check the model’s input controls in Create. A model accepting more than one image does not necessarily use the second image as an ending frame.
Can I animate an illustration or AI-generated image?
You can start with an illustration or generated picture. Describe the intended movement and visual style. Review the result for changes to the artwork, especially fine lines and character details.
Can a photo become a talking video?
A motion prompt alone does not guarantee speech or synchronized lips. Audio support varies by model. For a speech-focused workflow, explore the separate Lip Sync tool and check its input requirements.
Should I use Image to Video or Video Templates?
Use Image to Video when you want to write the movement yourself and select a model. Explore Video Templates when you prefer to start from an existing concept or effect.
Does this page upload my image?
The upload button takes you to Create, where you choose and upload the image. This landing page lets you explore models and prepare a motion prompt before you start.
How can I make a longer project?
For a sequence of scenes, explore Movie Studio. For a continuation of a clip, explore Extend Video. Check the input requirements and available settings in each tool.
Can I download the generated video?
You can review and download your completed video in Create. If the motion is not what you wanted, adjust the prompt and try another take.