AI Video Prompts — Sora, Veo & Runway Templates

AI video prompts.

Templates for text-to-video — scene, subject, camera move, lighting and style. Tap to copy, then swap in your details.

⌕

Tap Copy on any prompt, paste it into your AI assistant, and replace the [bracketed] parts with your own details.

Scene & subject

Cinematic establishing shot

[Subject] in [location], [time of day]. Cinematic establishing shot, [camera move, e.g. slow dolly in], natural lighting, shallow depth of field, photorealistic, 24fps film look.

Character action shot

[Character description] [doing a specific action] in [setting]. Medium shot, [camera move], [lighting], detailed and realistic motion, consistent character throughout the clip.

Product showcase

A [product] on [surface/background], rotating slowly, studio lighting, clean and minimal, soft reflections, macro detail, smooth camera orbit.

Camera & motion

Specify a camera move

[Your scene]. Camera: [dolly in / crane up / tracking shot / orbit / handheld]. Keep the motion smooth and deliberate, ending on [final framing].

Slow-motion moment

[Subject and action] captured in slow motion, high-frame-rate look, [lighting], emphasising [the detail or motion you want to highlight].

First-person POV shot

First-person POV of [activity], as if wearing a head-mounted camera, [environment], natural movement, immersive and realistic.

Style & look

Set a visual style

[Your scene]. Visual style: [e.g. 1970s film, anime, claymation, documentary, neon cyberpunk]. Apply it consistently to colour, lighting and texture across the whole clip.

Match a mood with lighting

[Your scene]. Mood: [calm / tense / dreamy / energetic]. Use lighting and colour grading to create it — [e.g. warm golden-hour light / cool blue shadows / high contrast].

Animated motion-graphic clip

A short motion-graphic animation of [concept], [colour palette], clean vector style, smooth transitions, looping, suitable for [use, e.g. an explainer intro].

Direct it like a shot list

Text-to-video tools like Sora, Veo, Runway and Pika respond best when you describe a clip the way a director would frame a single shot — not a whole story. Name the subject, the setting, one clear action, the camera move, and the lighting and style, then stop. Trying to cram a multi-scene narrative into one prompt usually produces something muddled; you'll get cleaner results generating short, specific clips and stitching them together. Keep each clip to a single continuous moment, and add style cues (film stock, era, palette) to lock the look.

FAQ

Which AI video tools do these prompts work with?

They're written as plain shot descriptions, so they work with Sora, Veo, Runway, Pika, Kling and similar tools. Some have extra controls (duration, aspect ratio) you can set on top.

Why does my AI video look glitchy or inconsistent?

Most current tools handle short, single-action clips far better than long or complex ones. Shorten the clip, simplify to one subject and one camera move, and describe the motion explicitly.