A vague prompt produces a generic video. A specific prompt produces a tailored, useful video. The difference is not magic — it is structure. Strong Runway prompts have five parts: subject, action, setting, style, and camera. Once you internalize this anatomy, every prompt you write becomes more effective.

A strong Runway prompt has five components. Subject tells Runway what to show ('a cat,' 'a city,' 'a person walking'). Action tells Runway what is happening ('playing with yarn,' 'at sunset,' 'through a forest'). Setting tells Runway where it happens ('in a cozy room,' 'on a beach,' 'in morning fog'). Style tells Runway the visual style ('cinematic,' 'documentary,' 'anime,' 'photorealistic'). Camera tells Runway the camera movement ('slow pan,' 'drone shot,' 'close-up,' 'static shot'). When all five are present, Runway has enough information to produce a tailored video.

Each component of the prompt narrows the space of possible videos Runway might generate. Subject narrows what is in the frame. Action narrows what is happening. Setting narrows the environment. Style narrows the visual treatment. Camera narrows the movement. With all five present, the output space is small enough that the video feels tailored. With any missing, Runway guesses — and guesses produce generic output.

1. Diagnose a vague prompt

2. Add subject and action

3. Add setting

4. Add style

5. Add camera

A user prompted Runway with 'A city.' The output was a generic cityscape with no particular character. He then rewrote with the five-part anatomy: 'Aerial drone shot over Tokyo at golden hour, cinematic, slow forward motion, photorealistic.' The second output was stunning — a specific, cinematic, beautiful video. Same model, same tool — the prompt structure made the difference.