Camera movement vocabulary that video models understand
The 14 camera moves that generation models respond to reliably, with the phrasing that works and the phrasing that produces a wobbling mess.
Use ONE of these per shot, written exactly like this: static shot, locked-off camera slow dolly in toward {subject} slow dolly out revealing {setting} slow push in on {detail} tracking shot following {subject} from behind tracking shot alongside {subject}, same speed handheld follow, slight natural shake slow pan left to right across {scene} tilt up from {ground detail} to {face / sky} slow orbit around {subject}, 90 degrees crane shot rising above {scene} drone shot, high altitude, slow forward rack focus from {foreground} to {background} whip pan to {new subject} Avoid: 'dynamic camera', 'epic movement', 'cinematic camera work', combined moves ('dolly in while orbiting and tilting').
Paste into the video model as one paragraph; remove the labels once you are happy with the content.
Steps
- Pick one move per clip. Combined moves are the single most common cause of warped geometry.
- Prefix with 'slow' unless you want speed; models default to too fast.
- Say where the move ends ('toward her face', 'revealing the harbour'): a move with a destination is drawn more steadily.
Tips
- A static locked-off shot plus subject motion is the most reliable output of every current model. Use it for anything that must look clean.
- For image-to-video, camera moves that reveal what was outside the frame (dolly out, pan) force the model to invent pixels; dolly in and rack focus keep it inside what it has.
| works with | Veo: Google text-to-video / image-to-video prompts. · Sora: OpenAI video generation prompts. · Kling: Kling text/image-to-video prompts. · Runway: Runway Gen prompts and camera controls. |
| cost | Free. The video or image model you paste it into has its own pricing. |
| status | Unverified: written from the technique, not yet run end to end by us. |
| license | CC BY 4.0 for the text here; skills are MIT. |
| updated | 2026-11-01 |
| for agents | camera-movement-vocabulary.md |
Related
The 7-part prompt structure for text-to-video
A fill-in template that gives any video model the seven things it needs: subject, action, setting, camera, light, style and sound. Fewer vague words, fewer wasted generations.
Image-to-video: generate the first frame, then animate it
Make a still with an image model first, fix the composition there, and only then ask the video model to animate it. Cheaper iterations and the look you chose, not the one the video model guessed.
Keep the same character across clips: the reference sheet prompt
A character description block you reuse verbatim in every prompt, plus a generated reference sheet, so a face, outfit and build stay the same from shot to shot.
Hook for a generated video: what must be in frame 1 when a model draws it
The first-frame rules for Veo, Kling and Runway: start mid-action, name the motion already happening, keep text out of the generation, and generate the hook shot as its own 4-second prompt first.
Product hero shot: a 5-second generated clip for an ad or a listing
Prompt for a clean product shot with one motion and one light change, the kind that sits at the top of a listing or opens a 15-second ad.
Anime style: 2D cel look, limited animation, clean line weight
A filled 7-part prompt for a hand-drawn anime shot: flat cel shading, animation on twos, a held painted background, and the three things the style breaks in current video models.