guides

Runway Gen-3 Video Prompts: Complete Guide to AI Video Generation

Understanding Runway Gen-3 Prompts

Runway Gen-3 generates short video clips from text descriptions. Unlike image generators, video prompts must describe motion, camera movement, and temporal changes in addition to the scene itself.

The Runway Prompt Formula

Scene: [description]. Camera: [movement]. Lighting: [type]. Motion: [what moves]. Duration: 5 seconds.

Camera Movement Keywords

  • Static: locked shot, tripod, fixed angle
  • Pan: slow pan left/right, sweeping pan
  • Tilt: tilt up/down, crane shot
  • Dolly: dolly in/out, push in, pull back
  • Tracking: tracking shot, follow shot
  • Orbit: 360 orbit, rotating around subject

Example Prompts

Scene: A lone astronaut walks across a vast Martian desert, red dust swirling around their boots. Camera: Slow tracking shot following from behind at waist height. Lighting: Golden hour with long shadows. Motion: Dust particles floating, cape billowing in wind. Duration: 5 seconds.

Tips

  • Always specify camera movement explicitly
  • Describe what is moving in the scene, not just the static elements
  • Use PromptFrom to generate video prompts from still images automatically

Related articles

Ready to try PromptFrom?

Upload any image and get AI prompts instantly.

Get started free