Create AI Videos from Text, Images, or Video
Create a short video from a written idea, animate a photo, or use an existing clip as a reference. Explore text to video, image to video, and video to video in one workspace. Add your own motion prompt or start with a community template to try a video trend.
How to Generate AI Videos
Start with text, an image, or a video
Write a prompt for text to video, upload a first frame in keyframe mode for image to video, or add a reference clip in multi-reference mode for video to video. Community templates can also load a prompt, settings, and references.
Describe the motion and choose settings
Explain how the subject and camera should move. Select the model, duration, aspect ratio, resolution, and available audio settings. Check the displayed credit cost before generating your clip.
Preview, refine, and download
Watch the full clip to check motion, subject details, and sound. Adjust the prompt or references if needed, then download the version you like. Add captions or combine clips in your usual editing app.
AI Video Trend Maker
Photo animation and moving portraits
Turn a selfie, pet photo, or illustrated character into a short animation. Describe a smile, a head turn, or a gentle breeze. Start with a small movement and review the subject’s details throughout the clip.
Transformations and creative video effects
Try a change in appearance, scale, or setting with a motion prompt. Explore community video templates for ideas, add your own subject and references, and describe how the change should unfold.
Camera moves and cinematic reveals
Try a slow push-in, an orbit around a subject, or a pull-back that reveals a larger scene. Guide the movement with a prompt or reference clip, and choose a supported format for TikTok, Reels, or Shorts.
AI Video Generator FAQ
Yes. In multi-reference mode, enter your prompt without uploading reference media. Describe the scene, action, and camera movement, choose the available settings, and generate a text to video clip.
Choose keyframe mode and upload your photo as the first frame. Describe the motion you want, then select the video settings and generate. You can also send a finished image from the image workspace using Generate video.
In multi-reference mode, upload a reference video and describe the new clip you want. The reference guides generation rather than being edited directly. How closely the result follows its motion and style depends on the model, prompt, and source clip.
Yes. In keyframe mode, upload a first frame and an optional last frame, then describe the movement between them. Check the selected model’s available formats and review how the generated clip transitions between the images.
Open a community video template, preview its result, and choose Use this template to load the prompt, settings, and references. Add your own materials and adjust the action or scene before generating your version.
Yes. The selected model determines the available durations, aspect ratios, resolutions, and audio options. Choose 9:16 for a vertical clip when supported, and check the displayed credit cost before generating.