Skip to main content

Image & Video Generation FAQ

1. How do I generate images using my model?​

Once you've successfully created your AI character model, head to the Character Studio or the main Create Page. Select your desired character. Then, in the prompt field, describe the scene, action, style, or attire you envision for your character. Our advanced AI will then generate an image based on your text input and character model.

2. Can I change my model's pose, clothing, or background?​

Absolutely! You have significant control through prompting. Clearly describe the desired pose (e.g., "sitting on a throne," "walking in a forest"), clothing ("wearing a futuristic silver jacket"), and background details ("neon-lit cyberpunk city at night") in your prompt. The more specific your prompt, the better our AI can match your vision. You can also use pre-defined or custom-uploaded poses if available for your character.

3. How do I create videos with my model?​

After generating an image you're happy with using your character model, you'll often find a "Convert to Video" option. This allows you to animate that specific scene. Alternatively, you can navigate to the "Image-to-Video" section of our Video Creator, upload any image (including one of your character), and apply various animation and motion effects.

4. What does 'image-to-video' mean?​

Image-to-Video is a powerful feature that allows you to take any static image—whether it's one you generated on KreatorFlow, a photograph, or other artwork—and bring it to life. Our AI analyzes the image and applies dynamic motion, camera movements, and effects to transform it into a short video clip, adding a new dimension to your visuals.

5. What are the best practices for face swapping?​

For optimal face swapping results, use high-quality, clear, and well-lit source (the face to copy) and target (the image/video to apply the face to) media. Ensure faces are relatively front-facing and not heavily obscured by hair, hands, or shadows. The higher the quality and clarity of the input faces, the more seamless and realistic the final swap will appear.