Midjourney Cinematic Prompts: Film Stills and 22 Examples

The cinematic prompt formula

"Cinematic" on its own is one of the most overused words in Midjourney prompts, and it mostly adds a vague moody gloss. A frame from a film looks the way it does because of specific choices: the shot size, the lens, the light, the colour grade and the shape of the frame. Name those and you get a still that feels like it belongs to a story.

1. Moment: a subject doing something, mid-action, with a hint of narrative.
2. Setting and time: place, era, time of day, weather.
3. Shot and framing: wide establishing shot, medium close-up, over-the-shoulder, low angle.
4. Lens and film texture: anamorphic, shallow depth of field, 35mm film grain, halation.
5. Light and grade: motivated practical light, hard backlight, teal and orange, bleach bypass, desaturated.
6. Parameters: a widescreen aspect ratio, often --raw.

Template: film still, [moment], [setting and time], [shot and framing], [lens and texture], [light and grade] --ar 239:100 --raw. Starting with "film still" tells Midjourney you want a frame, not a poster or a painting. For more on framing, see the guides to shot types and camera angles.

Describe looks, not people. Instead of naming a living director, describe what you admire in their frames: symmetrical pastel compositions, neon-soaked rain, long static wide shots, handheld grit. The description is more precise and keeps your prompt original.

Aspect ratios and colour grades

The aspect ratio does a lot of the cinematic work. Midjourney’s --ar takes whole numbers, so cinema formats are written as integer ratios.

Colour grade words that Midjourney responds to reliably:

• teal and orange: warm skin against cool shadows, the blockbuster look.
• bleach bypass: desaturated, high contrast, silvery and harsh.
• muted pastel grade: soft, whimsical, low contrast.
• neon noir: deep blacks with magenta and cyan practical lights.
• warm sepia, faded 1970s film: nostalgia, lifted blacks, soft grain.
• black and white, high contrast: classic noir shadows and hard light.

22 copy-ready cinematic prompts

Drama and thriller

• film still, a woman reading a letter at a kitchen table at dawn, cold blue window light, warm tungsten lamp behind her, medium close-up, shallow depth of field, 35mm film grain --ar 239:100 --raw
• film still, detective standing under a flickering streetlight in heavy rain, low angle, hard backlight through the rain, neon noir grade, anamorphic lens flare --ar 239:100
• film still, two strangers sharing an umbrella at a bus stop, over-the-shoulder shot, city lights blurred into bokeh, muted teal and amber grade --ar 37:20
• film still, an empty hospital corridor at night, one figure walking away, symmetrical one-point perspective, sickly green fluorescent light --ar 239:100 --raw
• film still, interrogation room, single overhead lamp, cigarette smoke curling, extreme close-up of tired eyes, high contrast black and white --ar 37:20

Epic and science fiction

• film still, lone rider crossing a vast salt flat at sunset, extreme wide establishing shot, tiny figure, long shadows, warm desert grade --ar 239:100
• film still, astronaut removing a helmet inside a cramped spaceship cockpit, practical panel lights on her face, handheld feel, teal and orange grade --ar 239:100 --raw
• film still, colossal ruined statue half-buried in a jungle, explorers in the foreground for scale, morning haze, volumetric light shafts --ar 239:100
• film still, futuristic city market at night in the rain, crowds with glowing umbrellas, neon reflections on wet ground, anamorphic flares --ar 239:100
• film still, medieval army waiting on a foggy ridge before battle, rows of spears, desaturated bleach bypass grade, wide shot --ar 239:100

Romance and coming of age

• film still, teenagers riding bicycles down a country road in late summer, golden hour backlight, lens flare, faded 1970s film colours --ar 37:20
• film still, couple dancing alone in a closed diner, jukebox glow, pastel pink and mint grade, medium wide shot, soft grain --ar 37:20
• film still, a boy watching fireworks from a rooftop, silhouette against the coloured sky, wide shot, warm nostalgic grade --ar 239:100
• film still, grandmother and granddaughter making dumplings in a small kitchen, steam in window light, intimate close framing, warm tones --ar 4:3 --raw
• film still, woman on a night train looking at her reflection in the window, passing lights streaking, melancholy blue grade --ar 239:100

Period and stylised

• film still, 1920s jazz club, singer under a single spotlight, smoke haze, high contrast black and white, deep shadows --ar 4:3
• film still, symmetrical pastel hotel lobby, bellboy standing perfectly centred, flat frontal composition, soft even light --ar 37:20
• film still, 1980s suburban living room lit by a television screen, kids in pyjamas, VHS-era colour, soft grain --ar 4:3
• film still, Victorian street at night in thick fog, gas lamps, horse-drawn carriage silhouette, cool desaturated grade --ar 239:100
• film still, Western saloon doors swinging open, gunslinger silhouetted against harsh midday sun, dusty air, warm sepia grade --ar 239:100
• film still, samurai kneeling in a bamboo forest in rain, wide shot, deep green tones, misty depth --ar 239:100 --raw
• vertical film still for a short video cover, dancer mid-spin on a wet street at night, motion blur, magenta and cyan light --ar 9:16

Tips and common mistakes

• Motivate the light. Mention where light comes from (a lamp, a window, a screen). Motivated light reads as cinema; random glow reads as illustration.
• Imply a story. A character mid-action or looking at something off-screen feels like a frame from a longer sequence.
• Use --raw for realism. It cuts Midjourney’s default gloss, which helps stills look photographed rather than painted.
• Avoid poster language. Words like poster, title or key art invite text and collage layouts. Midjourney also struggles with long exact text, so add titles later.
• Build a sequence. Reuse the same seed or a style reference to keep grade and texture consistent across several frames of the same "film".
• Do not use real actors. Describe original characters by age, clothing and expression instead.

Doing this faster with Dreamdrive

Dreamdrive keeps the full prompt with every capture, --sref code included, so a look you liked is a search away in your Library.