30 Midjourney-Style Prompts (and Where to Run Them).

The Midjourney prompts that consistently land all follow one structure: subject + style + lighting + composition + mood. Below are 30 copy-paste prompts built on that structure and grouped by look. Because Midjourney has no open API, you cannot run it inside a multi-model studio, so each section notes which getvivix image model gives the closest result. Paste any prompt into the getvivix image generator, swap the bracketed parts, and iterate.

One thing up front: these are written in plain English, not Midjourney shorthand. Flags like --ar 16:9 or --stylize 750 only work inside Midjourney. On other models you set aspect ratio in the panel and write the style intent into the prompt instead. Everything below is portable.

The prompt formula

Every prompt in this pack is the same five parts in order. Steal the skeleton, swap the nouns.

1. Subject— what the image is of, as concretely as you can say it.
2. Style— cinematic, oil painting, anime, 3D render, watercolor, and so on.
3. Lighting— golden hour, neon, soft window light, harsh noon, moonlit.
4. Composition— wide shot, close-up, top-down, rule of thirds, shallow depth of field.
5. Mood— serene, ominous, nostalgic, energetic, dreamlike.

Concrete adjectives beat clever sentences. "Misty pine forest at dawn" outperforms "a beautiful nature scene" every time, on every model.

Cinematic & photoreal

Closest model: FLUX 2 Pro or Imagen 4. Both hold detail under zoom and read prompts literally.

1. A lone lighthouse on a cliff at dusk, cinematic wide shot, golden backlight, moody storm clouds rolling in, film grain, dramatic mood.
2. Rain-soaked Tokyo alley at night, cinematic style, neon signs reflecting in puddles, shallow depth of field, lonely atmosphere.
3. An astronaut floating above a glowing blue planet, photoreal, hard rim light from the sun, centered composition, awe and silence.
4. A vintage diner on an empty desert highway, cinematic, late-afternoon warm light, low wide angle, nostalgic and still.
5. Close-up portrait of an old fisherman, photoreal, soft window light, weathered skin texture, shallow focus, quiet dignity.
6. A foggy mountain lake at sunrise, cinematic landscape, soft pastel light through mist, mirror reflection, serene mood.

Fantasy & concept art

Closest model: FLUX 2 Pro or Seedream for rich, painterly worlds.

7. A floating island city with waterfalls pouring into clouds, epic fantasy concept art, golden-hour light, sweeping wide shot, wondrous mood.
8. An ancient dragon curled around a crumbling stone tower, dark fantasy illustration, moonlit, low dramatic angle, ominous atmosphere.
9. A glowing enchanted forest at night, fantasy art, bioluminescent mushrooms and fireflies, soft volumetric light, magical and calm.
10. A weary knight resting against a giant fallen statue, concept art, overcast diffused light, rule-of-thirds composition, melancholy mood.
11. A desert nomad riding a sand-whale across dunes, imaginative concept art, harsh noon sun, extreme wide shot, adventurous feel.
12. A witch's cottage built into a hollow tree, storybook fantasy style, warm lantern glow, cozy framing, whimsical and safe.

Portraits & characters

Closest model: FLUX 2 Pro or GPT Image for faces and consistent character looks.

13. Portrait of a cyberpunk hacker, neon-lit, magenta and teal rim light, tight close-up, confident and intense mood.
14. A 1920s jazz singer at a microphone, vintage film style, warm spotlight, medium shot, smoky and glamorous atmosphere.
15. A young botanist surrounded by hanging plants, soft natural light, eye-level medium shot, curious and gentle mood.
16. A grizzled space mercenary in worn armor, photoreal portrait, cold blue light from a console, three-quarter angle, guarded look.
17. A child blowing dandelion seeds in a sunlit meadow, dreamy style, golden-hour backlight, low angle, joyful and nostalgic.
18. A regal queen in an obsidian throne room, oil-painting style, dramatic side light, symmetrical composition, powerful and cold.

Anime & illustration

Closest model: Qwen-Image or Seedream; for a pure anime look, our anime generator is tuned for it.

19. A girl watching shooting stars from a rooftop, anime style, deep blue night palette, wide establishing shot, peaceful and longing mood.
20. A ramen shop on a rainy street, cozy anime illustration, warm interior glow against cool rain, medium shot, comforting feel.
21. A samurai standing in falling cherry blossoms, anime art, soft pink dawn light, dynamic low angle, calm before action.
22. A mecha pilot in a cockpit, sci-fi anime style, glowing instrument panels, tight framing, tense and focused mood.
23. A small dragon napping on a stack of books, kawaii illustration, soft afternoon light, centered composition, cute and cozy.

Product, posters & type

Closest model: Ideogram 3.0 or Recraft V4— both beat Midjourney at readable text and clean layout.

24. A minimalist coffee bag on a marble counter, product photography, soft diffused studio light, top-down flat lay, clean and premium mood.
25. A retro travel poster of a mountain range with the text "EXPLORE," flat mid-century illustration, warm sunset palette, bold balanced layout.
26. A sleek perfume bottle on a black reflective surface, luxury product shot, dramatic single spotlight, centered, elegant and dark.
27. A bold gym event poster with the headline "RISE EARLY," high-contrast graphic style, hard directional light, energetic layout.
28. A pastel skincare set on linen, soft commercial photography, bright airy daylight, gentle shallow focus, fresh and calm mood.

Surreal & abstract

Closest model: FLUX 2 Pro or Nano Banana for clean edits and dreamlike scenes.

29. A staircase that loops into the sky, surreal art, soft pastel dawn light, symmetrical wide shot, dreamlike and disorienting mood.
30. A whale swimming through a city of clouds, surreal illustration, golden volumetric light, sweeping composition, peaceful and impossible.

Which model to reach for

The same prompt produces a different image on each model, which is exactly why running a few is the smart move. Here is the short version.

Why these run on getvivix, not Midjourney

Midjourney is a closed app. It runs through its web interface and Discord, with no general API in 2026, so no other studio can host it. That is the whole reason this pack points at FLUX 2 Pro, Imagen 4, Ideogram 3.0, Recraft V4, and the rest. Those models were built API-first, and getvivix runs more than 100 of them in one place on a single subscription.

The practical payoff: paste one prompt, fan it across three models, and keep the best frame for the price of a handful of credits. The exact credit cost shows on the model card before every generation, paid plans include a commercial-use license, and the free tier (30 credits on signup plus 30 dropped daily, no card) is enough to find your look. If you are weighing the two directly, the vivix vs Midjourney comparison lays out the trade-offs.

Frequently asked

What makes a good Midjourney prompt?

A good Midjourney prompt names the subject, the style, the lighting, the composition, and the mood in plain words, roughly in that order. For example: "a lone lighthouse on a cliff at dusk, cinematic wide shot, golden backlight, moody storm clouds, film grain." Stacking concrete adjectives beats one long vague sentence, and the same structure works on FLUX, Imagen, and Ideogram too.

Can I run Midjourney prompts on other models?

Yes. The prompts in this pack are written in plain English, so they run on any modern image model. Midjourney does not have an open API, so you cannot host it in a multi-model studio. On getvivix you paste the same prompt into FLUX 2 Pro, Imagen 4, or Ideogram 3.0 and get a comparable artistic result with the cost shown before you generate.

Does Midjourney have an API?

Not a broad public one. In 2026 Midjourney still runs mainly through its web app and Discord, with no general API that lets other tools host it. That is why platforms like getvivix run FLUX, Imagen, Ideogram, Recraft, and similar models instead, all of which were built API-first.

What do --ar and --stylize mean, and do they work elsewhere?

Those are Midjourney-only flags. --ar sets aspect ratio and --stylize controls how heavily the model applies its house art style. Other models ignore the flags, but you get the same effect by picking an aspect ratio in the panel and adding style words like "highly stylized" or "photoreal" to the prompt text.

Which model gives the closest look to Midjourney?

For Midjourney's painterly, moody default, FLUX 2 Pro and Imagen 4 get closest with a few style words added. For clean text, posters, and logos, Ideogram 3.0 or Recraft V4 beat Midjourney. The honest move is to run the same prompt across two or three and keep the best frame.

How many credits does one image cost on getvivix?

It depends on the model and resolution, and the exact credit cost is shown on the model card before you click generate, so there are no surprises. The free tier gives 30 credits on signup plus 30 dropped daily with no card, which is enough to test a prompt across several models before committing.

Run these prompts free on getvivix — or browse the full model catalog to pick your look. 30 credits on signup, no card, cost shown before every generation. Check pricing when you are ready to scale.