Runway Prompt Guide: 10 Copy-Paste Prompts for Gen-Model Cinematic Video

Runway was one of the first AI video tools to break out of the demo-reel phase and into real production — its Gen-series models are a staple in ad agencies, film prep and music-video workflows, largely because of one thing: director-level control. Camera language, reference images, region-level motion — Runway treats a prompt like a shot brief, not a wish. This Runway prompt guide covers how the Gen-series models read text, the formula that consistently lands, and 10 copy-paste prompts across narrative, product, abstract, documentary and animation work.

What Runway’s Gen-Series Actually Gives You

A quick verified snapshot before you prompt (Runway ships updates fast, so this is the current picture, not gospel):

• Gen-series models (Gen-4, plus the newer Gen-4.5 tier) for text-to-video and image-to-video. Output runs in short takes — think 5–10 second clips, built for inserts, ads, concept shots and b-roll, not full films.
• Reference-image workflows — feed character, object or location references so a subject stays consistent across shots.
• Camera Control — structured camera moves you apply as settings, instead of hoping the text encoder interprets “cinematic drift” correctly.
• Motion Brush — paint the region that should move and freeze the rest: steam off a cup, a flag in the wind, one performer in a still crowd.
• Act-One — performance capture that transfers an actor’s facial performance onto a generated character.
• Keyframes — set start and end frames and let the model interpolate between them.

Why Cinematography Vocabulary Is the Leverage

Runway’s prompting guides lean hard on filmmaking language — and the model rewards it. “Dolly-in,” “35mm,” “shallow depth of field,” “practical lighting,” “golden hour,” “crane up” are not decoration; they are the control surface. Two other traits of Gen-series prompting worth internalizing:

1. Direct register, not conversation. Runway explicitly warns against conversational phrasing (“please make the camera slowly move…”). Write commands and descriptions: “slow push-in toward the subject’s face.”
2. Positive phrasing only. Negative prompts are officially unsupported in Gen-series models — negative phrasing can produce the opposite of what you asked. Describe the state you want.

The formula that works:

[Shot type] of [subject + wardrobe/material texture], [action with beats],
[environment + time of day], [camera move], [lighting], [style/grade], [mood].

Three rules for Runway prompts:

1. Camera first, one move per shot. Name the move and its speed — “slow orbit right,” “static frame,” “handheld follow.” Two moves in one prompt means chaos in both.
2. Let motion words do the timing. “Steam curls upward,” “fabric caught mid-turn,” “light shifts slowly across the wall” — visible change over time beats a static mood list.
3. In image-to-video, prompt motion only. The image already defines the look; describe what moves and how. 10–30 focused words often outperform a long paragraph.

10 Copy-Paste Runway Prompts

Cinematic Narrative

1. The Doorway Reveal

Medium shot of a woman in a wool coat standing in a dark apartment doorway, warm
hallway light spilling past her onto the floor, she exhales and steps inside, coat
rustling, camera slowly pushes in past her shoulder, practical tungsten lighting,
muted cinematic grade, quiet bittersweet mood.

2. The Rain-Soaked Goodbye (two-shot)

Wide two-shot of two people facing each other under a black umbrella on a rain-slick
street at night, neon signs blurring behind them, one reaches out and lets go,
raindrops catching the light, slow lateral dolly, shallow depth of field, 35mm film
look, cool blue and magenta palette, restrained melancholy mood.

Commercial Product

3. The Fragrance Hero

Luxury product hero shot of a matte black perfume bottle on wet black stone, a single
water droplet rolls down the glass, condensation beading on the cap, camera slowly
orbits right at low angle, hard key light with soft rim light, deep shadows, glossy
highlights, premium commercial cinematography, sensual restrained mood.

4. The Watch Macro

Extreme macro of a steel dive watch on brushed slate, the second hand ticks forward,
light rakes across the rotating bezel and sapphire crystal, a dust particle drifts
through the beam, static frame with slow focus pull from crown to dial, cool studio
lighting, precision-engineering mood, crisp commercial grade.

Abstract Art & Fluid

5. The Ink Bloom

Macro shot of black ink blooming into still water against a cream background, silky
tendrils curling and dissolving in slow motion, tiny bubbles trapped in the swirls,
static camera, soft diffused studio light from above, high-contrast monochrome grade,
hypnotic meditative mood.

6. The Molten Glass Swirl

Close-up of molten amber glass swirling in slow motion, glowing ribbons folding over
each other, heat shimmer distorting the air, tiny sparks drifting upward, camera
slowly tilts up along the flow, dark studio background, dramatic firelight from the
material itself, rich orange and black palette, mesmerizing mood.

Nature Documentary

7. The Dawn Coastline

Aerial drone shot gliding forward over volcanic sea cliffs at sunrise, low mist
hugging the basalt rock faces, white waves exploding against the shore, a lone
lighthouse at the headland, golden hour light raking across the ridges, naturalistic
documentary grade, serene epic mood.

8. The Kelp Forest

Underwater wide shot of a sunlit kelp forest, tall green fronds swaying in the
current, silver fish threading through the canopy, god rays cutting through the blue
water, camera drifts slowly forward between the kelp columns, natural underwater
lighting, National Geographic documentary style, tranquil mood.

Stylized Animation

9. The Claymation Breakfast

Charming claymation scene: a tiny clay chef with rosy cheeks kneads dough on a
rustic wooden table, flour poofs into the air, an oversized jam jar topples and
spills in slow motion, warm morning window light, macro lens, stop-motion texture
with visible fingerprints in the clay, playful cozy mood.

10. The Watercolor Wind

Hand-painted watercolor animation style: a lone cyclist rides through a golden wheat
field at dusk, wind waves rolling across the grain, birds lifting from the stalks,
paper texture visible in the sky washes, slow tracking shot alongside the cyclist,
soft pastel palette, dreamy nostalgic mood.

Runway-Specific Tips the Pros Actually Use

1. Motion Brush beats prompt gymnastics. If only one element should move (steam, hair, water), paint it and freeze the rest — control no text prompt can express.
2. Use Camera Control for moves, text for content. Apply the camera move as a setting and spend your prompt words on subject, action and light. This is what Runway’s own guidance points toward.
3. Lock characters with references, not adjectives. “Consistent oval face” helps; the same reference image across three shots helps more.
4. Test cheap, then scale. Short clips and draft passes first — credits are real money, and a bad prompt burns them at the same rate as a good one.
5. For performance shots, look at Act-One. Driving a generated character with a real facial performance solves the “AI acting dead-eyed” problem faster than any adjective stack.

FAQ

What is the best prompt structure for Runway? Shot type + subject with material detail + action beats + environment and time of day + one camera move + lighting + style/grade + mood. Direct register, positive phrasing, cinematography terms.

Does Runway support negative prompts? No. Gen-series models don’t support negative prompts, and Runway’s docs warn that negative phrasing (“a man with no hair”) can produce the opposite result. Rewrite exclusions as positive states (“a bald man”).

How long can a Runway clip be? Generation runs in short takes — roughly 5–10 seconds per clip on the Gen-series models. Build sequences by generating shots and editing them together, and use keyframes to control where a shot starts and ends.

How do I keep a character consistent across shots? Use reference images for the character’s face and wardrobe, reuse the same references in every generation, and keep the camera framing similar (a face distorted in close-up stays distorted).

Which is better for Runway — text-to-video or image-to-video? Image-to-video when the look matters (product, character, key art): generate the still first, then prompt motion only. Text-to-video when you’re exploring shots and camera language from scratch.

Ten shots in, you’ll have a stack of cinematic clips — a good moment to let generated footage travel, pairing your edit with multi-platform distribution tools like FLOWNIB. And when you want to see how other creators are pushing Gen-series models, browse thousands of real prompts at VideosPrompt — watch the result, copy the prompt, make it yours.