Runway Prompts, Tools & Guides
AI video generator built for both text-to-video and image-to-video, with Gen-3/Gen-4 models. Image-to-video prompts should describe motion and camera movement, not appearance -- the starting image already defines what things look like. Works best on one continuous motion per clip, not a multi-beat story.
Runway Tools
all tools →
FAQ
What's the difference between text-to-video and image-to-video on Runway?
Text-to-video describes an entire scene from nothing. Image-to-video starts from an uploaded frame, so the prompt should describe what changes or moves, not what things look like -- the image already handles that.
How long should a single Runway clip's action be?
One continuous motion or moment works more reliably than a multi-beat story -- for a longer narrative, generate several short clips and combine them afterward rather than asking for it all in one generation.
Does camera movement need to be stated explicitly?
Yes -- naming the actual movement (a slow dolly in, a static shot) gives Runway something concrete to follow, rather than a vague mood descriptor like "cinematic."