What is reference to video AI?
Reference to video AI builds a new video around the images you provide, such as a character, a product, or a visual style. Instead of animating one photo, the model uses your references to keep the subject consistent while it creates a new scene from your prompt.
Simple process
How to make a video from references
1
Add your references
Upload images of the character, product, or style you want to keep. Some models also accept a reference video.
2
Write the scene
Describe the new scene and the action. Use the @ Reference button to point to a specific reference in your prompt.
3
Generate and review
Pick a model and settings, check the credit cost, and generate. Review the result in My Creations and download the clips you like.
Reference to video AI questions
Image to video animates the photo you upload. Reference to video creates a new scene and uses your images as a guide for how the subject or style should look.
Clear images with one main subject, good lighting, and little clutter. Add several angles of the same character or product when you need it to stay consistent.
Some models accept a reference video as well as images. If the selected model does not support it, the workspace asks you to remove the video or choose another model.
Each generation uses credits. The cost depends on the model and settings you choose, and the workspace shows it before you generate. See the Pricing page for plans and credit packs.
Credits are returned for eligible failed generations. A finished result that looks different from what you imagined is not a failed generation, so check your images and prompt before you start.
More AI tools
All AI toolsAI Video Generator
Animate a photo or turn a text prompt into a video with MiaReel.
Frames to Video AI
Create a short video from a start frame and an end frame in MiaReel.
Image to Image AI
Transform one or more uploaded images with a prompt in MiaReel.
Text to Image AI
Create images from natural-language prompts with MiaReel.
AI Image Upscaler
Upload an image, choose an output size, and make it larger and clearer in one step.