The Complete AI Camera Prompt Guide: 77 Shots and Movements
A practical guide to 77 camera angles, shot sizes, and movements for more AI video.
Ask an AI video generator for a "cinematic shot" or "dynamic camera movement" and you leave a lot to interpret. Ask for a slow dolly in, clockwise orbit, low-angle shot, or extreme close-up, and you give it a much clearer instruction.
That is the purpose of this guide: a practical library of AI camera prompts for 46 camera movements, 22 camera angles, and 9 shot sizes. Each one gives you the filmmaking term, what it does, and a prompt you can copy into an video model.
Table of Contents:
- What Are AI Camera Angles, Movements, and Shot Sizes?
- How to Write AI Camera Prompts
- 46 AI Camera Movements and Prompts
- 22 AI Camera Angles and Prompts
- 9 AI Shot Sizes and Framing Prompts
- Frequently Asked Questions
What Are AI Camera Angles, Movements, and Shot Sizes?
AI camera prompts use filmmaking terms to tell a video model where the camera is positioned, how the scene is framed, and how the camera should move.
There are three main controls to understand:
- Camera movement describes how the camera moves during the shot, such as panning, tilting, tracking, orbiting, or dollying.
- Camera angle describes where the camera is positioned relative to the subject, such as low angle, eye level, overhead, or profile.
- Shot size describes how much of the subject appears in frame, from an extreme close-up to an extreme wide shot.
You can use each one independently or combine them in the same prompt. For example:
low-angle medium close-up with a slow dolly in
You do not need to stack lots of camera terminology into every prompt. The goal is to give the AI a clear, physically understandable shot direction that matches what you want the viewer to see.
How to Write AI Camera Prompts
Every camera example in this guide was built using the same basic prompt structure.
Keeping that structure consistent makes it easier to change the camera without accidentally changing the subject, scene, or overall visual style at the same time.
1. Tell the AI What Your Reference Image Shows
If you're using a reference image to keep the character across consistent shots, don't just attach the image. Tell the model what it represents:
[Image 1] is the man.
2. Describe the Scene Before the Camera Movement
Next, write one sentence describing where the subject is, what they are doing, and what the environment looks like. Leave the camera instruction out for now.
Here are three scene descriptions used for the examples in this guide:
He stands alone in the centre of a vast empty concert hall, single overhead spotlight, red seats all around, wearing a black suit.
He stands at the bottom of a deep spiral concrete stairwell, seen from high above, cold overhead light, wearing a black hoodie.
He is seen from outside through a rain-streaked window, working at a desk in a warm lamplit room beyond the glass, wearing a black tee.
3. Keep Your Visual Style Consistent Across Shots
The third part is a short description of the visual treatment. For this guide, the same line was reused across the camera tests:
Cinematic 35mm anamorphic, shallow depth of field, filmic grain, natural colour.
Write the look once, then reuse it while you test different camera shots. If you're generating a multi-shot sequence, changing the style can make the finished edit feel like footage from several unrelated productions.
4. Describe Exactly How the Camera Moves
The camera prompts in this guide generally define four things: what does the camera physically do? What should stay fixed while it moves? What does the movement contribute to the shot? Where should the camera finish?
This is what separates vague prompts from expert prompts. You could write:
The camera slowly circles around him.
But a more useful camera prompt describes the movement physically:
clockwise orbit. Camera: circles the subject at a fixed radius, smooth and continuous, travelling clockwise. Framing: the subject pinned centre while the whole world rotates behind them. Effect: gravity and importance; a moment crowned. End: completes its arc with the subject still centred.
Build an AI Camera Prompt
Use the AI Camera Prompt Builder below to put the formula into practice. Add your reference and scene description, set a consistent visual style, then choose a camera angle, shot size, and camera movement to assemble a complete prompt automatically.
Once the prompt looks right, copy it and use it for your next AI video generation.
46 AI Camera Movements and Prompts
Camera movement describes how the camera changes position or direction during a shot. It can be as simple as a slow pan across a room or as complex as orbiting around a subject, flying overhead, or passing through an object.
The library below includes 46 camera movements for AI video. The prompts are designed to work as the camera section of the formula above. Add it after your reference, scene description, and visual style.
The camera move library
Explore 46 camera movements with video examples. Hover or tap a clip to play it, then copy the camera prompt directly into your AI video workflow.
22 AI Camera Angles and Prompts
A camera angle describes where the camera is positioned in relation to the subject. Unlike camera movement, the camera does not necessarily travel anywhere. The library below includes 22 AI camera angles, each with an example and a copy-paste prompt you can adapt to your own scene.
When prompting an angle, it helps to describe the camera's physical position as well as using the filmmaking term. Instead of only writing:
low-angle shot
You can make the geometry clearer:
low-angle shot. Camera positioned below the subject's eyeline, pointing upward toward them.
The camera angle library
Explore 22 camera angles with visual examples. Copy any angle prompt and add it to your AI video prompt to control perspective, framing, and point of view.
9 AI Shot Sizes and Framing Prompts
Shot size describes how much of the subject appears in the frame. It determines whether the viewer is focused on a small detail, a person's expression, their full body, or the wider environment around them.
The library below includes 9 AI shot sizes, each with an example and a copy-paste prompt you can adapt to your own scene.
The shot size library
Explore 9 essential shot sizes with visual examples. Copy a framing prompt and combine it with a camera angle or movement to control exactly how your subject appears in frame.
Frequently Asked Questions
What are the most useful camera movements for AI video?
Some of the most useful AI camera movements are dolly in, dolly out, pan, tilt, tracking, orbit, truck, crane, zoom, and handheld movement.
The best movement depends on what the shot needs to communicate. A dolly in can draw attention toward a subject, tracking can follow action, an orbit can reveal a subject from different sides, and a crane shot can show scale.
What is the difference between a camera angle and a camera movement?
A camera angle describes where the camera is positioned relative to the subject. A camera movement describes how the camera changes position or direction during the shot.
For example, low angle is a camera angle, while dolly in is a camera movement. They can also be combined in the same prompt:
low-angle medium shot with a slow dolly in
What is the difference between shot size and camera angle?
Shot size controls how much of the subject appears in frame, while camera angle controls the position from which the subject is viewed.
For example, a close-up and a wide shot are shot sizes. Eye level, low angle, overhead, and profile are camera angles. You can combine both to create more specific framing, such as an overhead wide shot or a low close-up.
What is the difference between a dolly and a zoom?
A dolly physically moves the camera closer to or farther away from the subject. A zoom keeps the camera in the same position and changes the lens magnification.
Because a dolly changes the camera's position within the scene, foreground and background objects shift in perspective. A zoom changes how large the subject appears without creating the same perspective change.
If an AI model confuses the two, make the distinction explicit in the prompt:
Camera physically moves forward through the scene. No lens zoom.
How do I prompt a tracking shot in AI video?
Describe what the camera is following, which direction it travels, and what framing it should maintain. For example:
Side tracking shot. Camera travels alongside the subject at the same speed, maintaining a consistent medium-wide shot while the background moves past behind them.
Keeping the relationship between the camera and subject clear can help prevent the model from drifting away from the action.
How do I keep the camera still in an AI-generated video?
Ask for a locked-off static shot and explicitly state that the camera should not move. For example:
Locked-off static shot. Camera fixed in one position on a tripod, zero camera movement throughout the clip, identical framing from beginning to end.
Static shots are especially useful when the subject or action within the frame should provide the movement instead.
Can I combine multiple camera movements in one AI prompt?
Yes, but simpler prompts are generally easier to control.
Start with one primary camera movement and add another only when the two movements form a clear, physically understandable shot. Stacking several unrelated moves can give the AI conflicting instructions and make the result less predictable.