Knowledge base · video
Wan
Video generation from text, photos and a reference + talking video (speech-to-video), a universal video model.
Open in studio →What it can do
- ✓Text-to-video and photo animation (image-to-video).
- ✓Generation from a reference.
- ✓Speech-to-video: a talking frame to speech or audio.
How to write a prompt
Subject → action → camera movement → atmosphere. To animate a photo, attach the picture; for speech-to-video, audio or text.
A prompt that works
A girl walks through an autumn park, leaves swirl in the wind, the camera follows her, warm light, cinematic
Avoid this
“a video of a girl”
Describe the action and the camera movement, otherwise the frame is static and random.
Examples
Photo animation
(attach a photo) A light wind and motion, slow camera push-in
Variants within the group
- Wan 2.6 / 2.2 · T2V / I2V / Reference / Speech-to-Video, The fresh generation gives better motion, modes are chosen for the task.
Tips
- →One motion accent per clip.
- →The fresh generation for the final result.
FAQ
How much does a generation in Wan cost?
from 8 to 50 tokens per generation, depending on the chosen version and settings. Current plans and token packs are on the /en/pricing page.
Can the results of Wan be used commercially?
Yes, with an active Pixyn subscription. Note that some models have provider restrictions by industry, check the terms of the specific version.