CraftPrompts

AI video prompt generator

Video models think in shots. Choose the framing and camera move, describe what happens, and get a shot-list style prompt for Sora, Veo, Runway or Kling.

Video prompt
shotcamera movesubjectactionsettingstyleaudio
Fill in a few fields and your prompt is set here, line by line.
0 words · 0 characters

Direct it with AI

AI writes a richer, time-ordered shot description from your idea.

Camera language cheat sheet

TermWhat you get
Push in / pull outThe camera moves toward or away from the subject. Push-ins build tension; pull-outs reveal context.
Pan / tiltThe camera rotates horizontally or vertically from a fixed spot.
Tracking shotThe camera travels alongside a moving subject.
OrbitThe camera circles the subject, a popular “hero” move.
Crane upThe camera rises, often ending on a wide view.
HandheldSlight natural shake: documentary feel.

Start still images with the image prompt generator; many video tools can animate a still you like.

Questions

What makes a good AI video prompt?

Describe one continuous shot in present tense: the shot size and camera movement first, then the subject and what they do over time, the setting, lighting and style. OpenAI’s Sora guide and Google’s Veo guide both recommend specifying camera language explicitly.

Does this work for Sora, Veo, Runway and Kling?

Yes. All of them accept natural-language descriptions with film vocabulary. Veo and newer Sora versions also generate audio, so the optional audio line is used there and ignored elsewhere.

Why only one shot?

Current models produce short clips, typically 5 to 20 seconds. Asking for several cuts in one clip usually produces a muddle. Generate each shot separately and edit them together.

How do I keep a character consistent across clips?

Reuse the exact same subject description word for word in every prompt, and use the tool’s reference image or character feature if it has one.

More tools in the studio