AI Video Prompt Examples: Templates and a Real Breakdown

The single biggest lever on the quality of an AI video is the prompt. The tools are good enough now that the difference between a flat, generic clip and a genuinely cinematic one usually comes down to how you describe the shot. This post collects practical AI video prompt examples you can copy and adapt, breaks down a real prompt-to-video generation, and shows you the repeatable structure behind prompts that actually work.

What a good AI video prompt actually contains

Weak prompts describe a thing ("a man fishing on a lake"). Strong prompts describe a shot — the subject, what they do, how the camera behaves, the light, and the mood. Almost every high-quality AI video prompt covers these six ingredients:

• Subject — who or what is on screen, with specific visual detail (clothing, age, materials, colour).
• Action — the single clear thing that happens in the shot. One beat per clip beats five.
• Camera — angle and movement: over-the-shoulder, slow push-in, locked-off, drone pull-back.
• Lighting — time of day and quality of light: dawn, golden hour, hard side-light, moonlight.
• Mood & atmosphere — fog, dust, haze, reflections — the details that sell a scene.
• Style & framing — "photorealistic, cinematic 16:9", or a specific film look.

Keep it concrete and keep it to one moment. If you find yourself writing "and then, and then, and then," you probably want two clips, not one.

A real AI video prompt example, broken down

Here is a prompt we actually ran — you can watch the exact output further down this page. Notice how it hits every ingredient above without being bloated:

The scene opens at dawn on a tranquil lake, with soft fog hovering above the water's surface… A young man, dressed in a light green jacket, stands on a wooden pier, gripping his fishing rod tightly… The camera, positioned just over his shoulder, follows the movement fluidly to the right… culminating in a gentle push-in on the expanding ripples in the water.

Why it works:

• Subject is specific: "young man, light green jacket," not just "a person."
• Action is a single readable beat: he jerks back, the line goes slack.
• Camera is explicit: over-the-shoulder, follows right, then a gentle push-in.
• Light & mood carry the shot: dawn, warm yellow hue, drifting fog, shimmering reflections.

This was generated as an image-to-video clip (starting from a still frame, with a handheld motion style). Starting from an image gives you far more control over composition and character than text-to-video alone — the prompt then directs motion and camera rather than inventing the whole frame.

Copy-paste AI video prompt examples

Use these as templates — swap the bracketed parts for your own scene. Each one is a single, self-contained shot.

Cinematic establishing shot

Wide aerial establishing shot of [location] at golden hour. Slow drone push-in toward [subject]. Warm low sun, long shadows, light haze in the air. Photorealistic, cinematic 16:9, shallow depth of field.

Character close-up with emotion

Tight close-up of [character description], lit by soft window light. She slowly looks up and her expression shifts from worry to relief. Locked-off camera, gentle push-in. Photorealistic skin detail, shallow focus, muted colour palette.

Product hero shot

Studio shot of [product] on a matte surface, rotating slowly on a turntable. Clean gradient background, soft key light with a subtle rim light. Camera orbits 90 degrees left to right. Crisp reflections, premium commercial look, 16:9.

Action beat

Low-angle tracking shot of [subject] sprinting through [environment]. Handheld camera keeps pace, slight motion blur, dust kicked up underfoot. Hard side-light, high contrast, fast shutter feel. Cinematic, tense.

Nature / mood

Misty forest at dawn, shafts of light breaking through tall trees. Slow camera drift forward along a narrow path, fog rolling across the ground. Soft ambient light, deep greens, calm and atmospheric. Photorealistic 16:9.

Common mistakes that ruin AI video prompts

• Too many actions in one clip. Models lose coherence. Split it into shots.
• No camera direction. If you don't say how the camera moves, you get a static, lifeless frame.
• Vague subjects. "A car" gives you a random car. Describe make, era, colour, condition.
• Forgetting light. Lighting is 80% of the "cinematic" feeling — always specify it.
• Fighting the model with contradictions. "Fast chaotic action" plus "calm and still" confuses the generation.

From a good prompt to a finished film

Prompt writing is one skill in a longer pipeline — character consistency, scene planning, editing, sound, and stitching shots into a story are what turn clips into a film. If you want the complete, structured workflow instead of trial-and-error, that is exactly what we teach step by step in our AI video course. You will learn prompt engineering, image-to-video control, character consistency, and how to assemble everything into finished, publishable work.

Start with the templates above, run a few tests, and study what changes when you adjust one ingredient at a time — then take the AI filmmaking course to put the whole pipeline together.