Luvio AI · 光趣 AILuvio AI · 光趣 AI
AI video generation from text and images
Create video clips in Luvio AI using text or supported image references, with model, duration and visual settings for storyboards and creative scenes.
Describe the action, light and mood of one shot, then turn it into video material you can develop further.
Describe a shot in text
Text-to-video is useful for exploring a scene from scratch. Describe the subject, setting, action, camera movement and light. Keeping one main action in each request makes the result easier to assess than combining many scenes.
Bring reference material into generation
Image-to-video can start from a prepared visual and add movement. Support for first frames, last frames, reference images and reference videos varies by model. Upload only the material supported by the selected fields.
Continue on the canvas or in a project
Homepage video creation can continue on the canvas. Short drama projects organize episodes and storyboards. Check characters, actions and composition before revising prompts or references. A generated clip is source material, rather than a fully edited drama.
Getting started
Define the shot
Identify the subject, setting and action, and prepare references you have permission to use.
Choose a model and settings
Review available models and supported aspect ratios, resolutions, durations and modes. Check the displayed points before submitting.
Generate and review
Sign in, submit the request and inspect the clip. Keep usable versions for later storyboard or editing work.
Common questions
When should I use text or image references?
Use text to explore a new scene. When a character, composition or setting is already defined, use image references where the model supports them. Check the workspace for current capabilities.
Can I generate without signing in?
You can read these introductions publicly. Uploading, generating and saving creative work requires signing in. Points are shown before submission.