Appearance
Video Generation
Select a video model from the model dropdown, describe your scene, and the video is generated inline.
Note: Video generation uses significantly more points than text or image tasks.
Available Models
Sora (OpenAI)
| Model | Notes |
|---|---|
| Sora 2 Pro | Highest quality, cinematic |
| Sora 2 | Standard quality |
Good for: High-quality short films, product demos, cinematic sequences with synchronized dialogue
Veo (Google)
| Model | Notes |
|---|---|
| Veo 3.1 | Full quality, smooth motion |
| Veo 3.1 Fast | Faster, lower cost |
Good for: Realistic motion, nature scenes, physics-accurate content, narrative sequences
Seedance (ByteDance)
| Model | Notes |
|---|---|
| Seedance 2.0 | Latest, best quality |
| Seedance 2.0 Fast | Faster version |
Good for: Natural motion, character consistency, social content, multi-shot narratives
Other Models
| Model | Notes |
|---|---|
| Grok Video 1.5 | Fast, creative, native audio sync |
| Midjourney Video-V1 | Midjourney aesthetic |
Writing a Good Video Prompt
Describe: Scene · Camera movement · Mood
Example:
A golden retriever running through a field of sunflowers in slow motion, warm afternoon light, cinematic quality, static wide shot
Tips:
- Keep actions simple — complex multi-step actions reduce quality
- Specify camera:
slow pan left,static shot,drone aerial view - Focused, short scenes produce better results than vague descriptions