Choosing the right text prompt for Midjourney is not just vital but essential to generate the most jaw-dropping results in Midjouney the world has yet to be seen.
In this article, you’ll see 131 text prompts for Midjourney and their results.
For the remainder of the post the goal is simple, you’ll learn what a text prompt is and how to use a text to image prompt in Midjourney.
Make sure your sunglasses are within reach, the results can be blinding.
Prepare to be dazzled!
Feel free to use the Table of Contents to navigate ahead.
What is Midjourney AI?
Midjourney is a renowned research lab that specializes in both design and AI.
Through the Discord platform, their most popular resource – an unparalleled text-to-art AI generator – has revolutionized how we create or generate artwork today.
How to Gain ACCESS to Midjourney?
Here’s a quick step-by-step guide to getting access to Midjourney.
1. To take advantage of Midjourney’s services, you’ll need to set up a free Discord account. Don’t worry if you haven’t signed up for it yet; the process is quick and easy, click here to sign up for Discord.
2. Take a trip to the Midjourney website and select “Join the beta” to be directed directly to an exclusive Discord invite.
3. Click the Discord invite to Midjourney. Choose to Continue to Discord.
4. When you open the Discord app, look to your left and you’ll spot a ship-shaped icon that says “Midjourney”, click on it and get ready for an exciting adventure!
5. If you’re looking for a place to start your Midjourney experience, look for the “Newcomer” rooms.
From there, you can find several open channels with names such as “newbies-110”.
Note: These room numbers changes depending on the time you sign up.
Pick any of these newbie rooms and get ready to begin.
This is where you enter your prompt in Midjourney to generate your image.
What’s a prompt you asked? Let’s pin that for later in this blog.
You’re here for the 131 text prompt for Midjourney and I’m here to deliver.
131 Midjourney Prompts Examples
All these were generated within Midjourney and only using text to image prompts.
Somewhere in the list of images add one of the example this: As you can see we have two versions of the same prompt.
Note: Feel free to try these prompts yourself and see what you’ll create within Midjourney.
Using the same prompts in the below list will yield different within Midjourney. Also, the parameter –v 4 is not needed as the current version of Midjourney automatically adds it internally.
2D Illustration Art
/imagine style of Pierre-Joseph Redoute, different types of mushrooms, botanical illustration, plain background
/imagine oasis, with lake in the middle, mid day, desert, complex, wide view + color pen and ink, illustration, realistic, maximalist, spectacular details, 8K, concept art, cinematic, atmospheric, epic composition, dramatic light, + vibrant colors, high contrast –v 4 –ar 2:3
/imagine sticker, A super cute orange tabby kitten, vector
/imagine tshirt vector illustration, shiba synthwave, vivid colors, detailed
/imagine elemental animal spirit of paisley –v 4
/imagine flower valley surrounded by mountains. The colors bright and vibrant, a sense of beauty and adventure, vector illustration –ar 3:2
/imagine stunning vector art tango dancers, highly detailed, intricate, blue, orange and white
/imagine full page of illustrated spells from a Witch Grimoire by Jody Bergsma, ink outlines, watercolor splash background, intricate details
3D Concepts
/imagine Level 1 wizard, full body, blender 3d, Disney Pixar, Mobile game character, isometric, clash royale, cute
/imagine game sheet wizard items, clay render, blender 3d, style of 3D mobile game, isometric, clash royale
/imagine homogeneous, big city at night, new york, inside view, finite texture, micro decomposition, disharmonic entropy, limited field, simple chaos, land near scope, ancient natural biotope, crystal ocean, absorption, simple waterflow, symmetry ice, stained silver, future genotype, proton shield, unlimited power, 8k, retracing, hyperrealistic, super detailed photo –ar 3:2 –chaos 30 –v 4 –q 2
Abstract Art
/imagine musical notes, ethereal sky, majestic light, dreamy, vibrant color
/imagine fidenza color wave
/imagine alberto seveso art, red and yellow, water ink, ink water, ink cloud, alberto seveso art, loose painting style, intricate detail, cinematic lighting, octane render, 8k render, volumetric lighting
/imagine fidenza pigmentation swirl
Adorable Art
/imagine chibi, anime, genshin, illustration –v 4 (Version 1)
/imagine chibi, anime, genshin, illustration –v 4 (Version 2)
/imagine adorable cartoon tiny puppy shiba, out of a teacup, pastel enviroment –niji
Anime Art
/imagine lofi vibe, silouette of a girl from a distance, dancing to a busy city, raining –ar 3:2 –q 2
/imagine masterpiece, best quality, high quality, A very bright, sunny day, anime girl sitting on a glove compartment, in a panama hat, watercolor, very detailed, golden hair, hairstyle square, golden eyes, hands behind her back, anime style –q 2 –ar 3:2
/imagine the silhouette of a woman looking at a sunset, pink and purple clouds, anime –niji –ar 3:2
/imagine the most beautiful and sad japanese waifu, hurt, wet semi transparent outfit, 20-years old girl, smooth pale skin, looking at me, close-up portrait, “modern anime”:: skinny , DJRim , highly detailed, super high quality, key visual of anime, pixiv masterpiece, professional color grading, 8k –ar 3:2 –niji (Version 1)
/imagine the most beautiful and sad japanese waifu, hurt, wet semi transparent outfit, 20-years old girl, smooth pale skin, looking at me, close-up portrait, “modern anime”:: skinny , DJRim , highly detailed, super high quality, key visual of anime, pixiv masterpiece, professional color grading, 8k –ar 3:2 –niji (Version 2)
/imagine dallas city in the style of ghibli, beautiful city, clouds on blue sky, retro anime –niji
/imagine cherry blossom temple in cyberpunk illustration style
/imagine the great wave off Kanagawa, anime style –ar 3:2
Animal Art
/imagine adorable, cute pixar style baby orange tabby kitten, fluffy, focus on eyes, looking at camera, wearing a tuxedo, delicate and fine, high detailed, warm color, natural light, simple background, octane render, ultra wide angle, 8K, wildlife photography
/imagine adorable cute baby pixar style shiba, fluffy, focus on eyes, wearing a red bowtie, delicate and fine, high detailed, warm color, natural light, simple background, octane render, ultra wide angle, 8K, wildlife photography –v 4
/imagine spectral white fox sitting on a balcony of a gothic castle –q 2 –v 4
Architectural Art
/imagine organic house embedded into the greenland, surrounded by greenery, architectural photography, futurism, modernist architecture, concept design –q 2 (Version 1)
/imagine organic house embedded into the greenland, surrounded by greenery, architectural photography, futurism, modernist architecture, concept design –q 2 (Version 2)
/imagine wooden chalet, A shape, mountains, sunset, mountain, ocean, realistic
/imagine cabin in the mountains, vibrant sunset, watercolor anime stylle
Car Concepts Images
/imagine futuristic sedan car, different blues, commercial driving through water –ar 3:2 –q 2
/imagine EV car with glowing car lamp light and EV charger station –ar 3:2
/imagine EV car with glowing car lamp light and EV charger station –ar 3:2
/imagine high resolution photography of a futuristic retro car, high gloss polished surfaces, wide transparent windows, cinematic lighting –v 4 –q 2
/imagine Electric Hypercar trending on artstation, minimal design, cinema 4D render, photorealistic, ad campaign, driving off-road on rocks in beautiful landscape, hyper detailed –q 2 –chaos 50 (Version 1)
/imagine Electric Hypercar trending on artstation, minimal design, cinema 4D render, photorealistic, ad campaign, driving off-road on rocks in beautiful landscape, hyper detailed –q 2 (Version 2)
/imagine high resolution photography of a futuristic retro car, high gloss polished surfaces, wide transparent windows, cinematic lighting –v 4 –q 2
Comic Art
/imagine oblique-angle full-body candid-shot of female, comic art style, street fighter style –niji
Cyberpunk Art
/imagine Pixar, DJ, headphones, cyberpunk, flashing lights in a club –v 4
/imagine female sideview, cyberpunk apparel website background, neon, glow retro future –ar 3:2 –q 2
/imagine a despondant 20 something casually dressed female cyberpunk sitting in a window still, future cityscape out the window flying cars blimps lots of neon lights, dynamic lighting, shadows, ultra detailed, hyper realism 8k, –ar 3:2
/imagine mysterious supermodel anime geek DJ girl is playing in front of the nicht city penthouse panolama view stage, chill atmosphere, overwatch style, cinematic lighting, anime, ultra detail –ar 3:2 –q 2 –v 4
/imagine beautiful, website, ui/ux, ui, futuristic, viser glassess, metaverse hardware, technology, background lighting, neon cyberpunk, ominous, tenebroso, ultra realistic, ultra resolution, ultra detailed, hyper-detailed, highly detailed, cinematic lighting, 8k, unreal engine, elaborate
/imagine Cyberpunk, futuristic, super detail, cluttered room, flower plants, Surrounded flower, backlighting, Landscape Window, Front angle, –ar 3:2
Cinematographic Style Art
/imagine retro fututistic space station evil cyberpunk astronaut, style by stanley kubrick, ridley scott, george lucas, spacesuit by vivienne westwood, hyper detailed, volumetric lighting, cinematic, 8k –ar 2:3 –q 1 –v 4
/imagine person walking towards a fractal, hyper realistic, majestic light, vibrant colors, flowers
/imagine 2 men and one girl standing in a planet , full body , cyberpunk , epic background galaxy another palnet , from future ,epic detailed , full color illustrated , octane render , hyperrealistic , cinematic , blue lunar horizon , height resolution , 8k ,micro details , –v 4
1 stunning film scene :: the entrance way to the summerland midnight homecoming ball, intricate details, welcome to the mysterious summerland kingdom, endless flowing silks upper in the air, countless fairy lanterns shinning in the trees, starry midnight sky, Disney style, Fantastic Beasts style, extremely long distance perspective, wide open air view, gorgeous, like a dream, octane render, hyperdetailed, hyperrealism, fantastic realism, photorealism, great depth, vivid colors, 3d render, volumetric lighting, photograph, front lighting, cinematic lighting, fantastic, dreamful colors, high contrast, fantasy, fairytale, extremely detailed, high definition, UHD, 16k –v 4 –q 2
/imagine surreal concept art, a world without time, enter the void, free roaming souls, creation, vibration, futuristic accents, high contrast, digital art, digital glitch, lens flare
Double Exposure Art
/Imagine 3D double exposure, side profile of a beautiful women with mountains and forest within
Fantasy Art
/imagine Back view lady, magical orange butterflies, wearing blue chiffon dress, orange glitter everywhere, blue forest background, moonlight, bokeh, fairy light, high details, by Anna Dittmann, fantasty, trending on artstation, paint strokes textures, –v 4 –ar 2:3 –q 5
/imagine happy Julianne Moore moonlight fairy goddess on stars throne, moon, moonlight, full body, fairytale celestial sky, with glittering translucid wings, bokeh, fairy light, high details, by Anna Dittmann, fantasty, trending on artstation, paint strokes textures, –v 4 –ar 2:3 –q 5
/imagine fashionable queen of the monstera, symmetrical face, artgerm, flowing hair, portrait, muted colors, artstation, character concept art, border and embellishments inspiried by alphonse mucha, fractals in the background, galaxy, light bokeh –ar 3:2
/imagine holographic psychodelic outer space in pastel cold color::2 landscape of desert::3 roads far away::4 stars and galazies::5 photography, hyperrealism, high resolution super realistic hdr 8k
/imagine A colossal burning bioluminescent tree with red glowing fireflies in the center of a lake surrounded by snowy mountains under a star filled sky at dawn, misty foreground, realistic –ar 2:3 –q 2
/imagine photorealistic mythology fantasy inspired landscape image, intricate design, unreal engine 5 render –q 2 –ar 3:2 –v 4
/imagine synthwave concept art, purple horizon, shadow figures, dolphins, bioluminescent purple, photo-realistic, cinematic lighting, atmospheric, low contrast, digital art, lens flare
/imagine surreal, milkyway in a glass bottle, forest setting, night time, focus blur, 8K, unreal engine, octane render –q 2 (Version 1)
/imagine surreal, milkyway in a glass bottle, forest setting, night time, focus blur, 8K, unreal engine, octane render –q 2 (Version 2)
/imagine surrealistic painting of a dragon egg::5 :on the inside stormclouds over magical pine forest under a glass dome::3 :on the outside the eggshell is made of a starry nebula::4 :silver dark ratio::1 :dark background, octane render, 4k, hyperdetailed, photorealistic, diverse textures, exciting variety of surfaces, dramatic lighting –ar 2:3
/imagine weapons, duel swords, one fire, one ice, glowing, Silver and white medival handles, one flaming, one icing, dark background, high detailed
/imagine glaxay in a circle, stars, planets, moon, sun dancing in the dark blue sky many bright colors
/imagine character traveling towards the night, masterpiece, In the style of Beksinski and Leonid Afremov, starry Night, Pattern, mystical
/imagine tree of life on mars (Version 1)
/imagine tree of life on mars (Version 2)
/imagine tree of life on mars (Version 3)
Fashion Art
/imagine Beautiful high-heeled shoes, glowing floating luminous crystal stars and crystalline candy, Intricate, Elegant, Scenic, Hyper-Realistic, 8k resolution, maximalist intricate, 3d render, pastel color ethereal mythical fantasy (Version 1)
/imagine Beautiful high-heeled shoes, glowing floating luminous crystal stars and crystalline candy, Intricate, Elegant, Scenic, Hyper-Realistic, 8k resolution, maximalist intricate, 3d render, pastel color ethereal mythical fantasy (Version 2)
/imagine Stunning antique minimalist black and purple dress, black inside, External gradual change color transparent cloak, Slim, Embroidered with dragon and flowers,Translucent crystal, beautiful elegant and noble, Diamond setting, Elegant, Fairy spirit flying, Digital rendering, Dream, High definition, Details, Magic Octane Rendered Surreal Light Background, 8k
/imagine ying yang fashion, elegant gown, intricate details, exquisite elegance, made of ying yang, flowing gown
/imagine futuristic fashion wolverine as tuxedo, Costumes and props, Knolling, Knolling layout, Natural materials, Deconstruction, Highly detailed, Depth, Lumen render, 8k
/imagine futuristic fashion wolverine as dress, Costumes and props, Knolling, Knolling layout, Natural materials, Deconstruction, Highly detailed, Depth, Lumen render, 8k
Floral Art
/imagine golden spectral flower , with soft , fairy fog background , light smoke , light lame , on the rose , light golden rose , transparent petals , clear , lifelike , real petals , feeling like love ,, complex details processing , high definition ,, dream fairy tale kingdom high definition , urea , 8K hd –v 4 (Version 1)
/imagine golden spectral flower , with soft , fairy fog background , light smoke , light lame , on the rose , light golden rose , transparent petals , clear , lifelike , real petals , feeling like love ,, complex details processing , high definition ,, dream fairy tale kingdom high definition , urea , 8K hd –v 4 (Version 2)
/imagine ornate pattern of flowers and vines, gold and iridescent, Southern lights, purple, teal, soft pink, intricate pattern, maximum detail, UHD, 32k photorealistic, HDR, High octane render –q 2
/imagine watercolor pattern soft yellow, green and gold dhalias, feathers, delicate, elegant and ethereal, luxury, volumetric, octane render –v 4
/imagine flowers megenta,luxury Gradient, teal blue black, digital art, line work, minimalist, bald colors v4
Futuristic Style Concepts
/imagine street photography, side view, futuristic fashion line for cyberpunk suits, foreground blur
/imagine robotic cyborg woman with metal visor, vr headset , chrome sci-fi cyberpunk armor, cinematic photorealistic atmospheric
/imagine object photography. beautiful mechanism gundam, unreal engine, wide shot, 16k, neon blue
Gems and Jewelry Art
/imagine Holographic purple gemstone made from liquid dreams ar 3:2 –v 4
/imagine teal green crystal, oval, bio luminescent, shadows, reflective black background, 8k
/imagine light blue crystal, oval, bio luminescent, shadows, reflective dark blue black background, 8k
/imagine orange crystal, oval, bio luminescent, shadows, reflective black background, 8k
/imagine all 7 chakras as fine gems, on a glass reflective table, out fo focus, macro photography, dark background –q 2
Isometric Art
/imagine Isometric clean image of cherry blossom forest and a mountain
/imagine Isometric clean pixel art image cherry blossom with mountain
Illusion Art Concept
/imagine black and white Optical Illusion, confusing Pattern render, HD, 8K
Interior Design Art
/imagine interior deisgn, perspective of kitchen room, modern open lighting room, mostly glass, natural light, minimalistic design, real estate photography –q 2
/imagine ellegant seemless luxury wall themed pattern (Version 1)
/imagine ellegant seemless luxury wall themed pattern (Version 2)
Landscape Art
/imagine dark vibrant sky with aurora borealis, steep mountains surrounding a clear lake, glow lighting, super realistic, 8k resolution, psychedelic, perfection –ar 3:2
/imagine Midway pinball, intrincated retro pinball, glowing innerlight intricate detail, platforms, mazes and lights / glowing innerlight intricate details / fine shiny metallic edges / photorealistic / cinematic lighting –v 4 –q 2
/imagine big bend national park, sunset, cloudy, inside crystal ball, hyper realistic, –v 4
/imagine vector art, illustration, yosemite, highly detailed, vibrant –v 4
Lofi Vibes Art
/imagine neon, lofi, girl working on computer, listening to music, heyper detail (Version 1)
/imagine neon, lofi, girl working on computer, listening to music, heyper detail (Version 2)
/imagine a hyper realistic 4k photo of a futuristic apartment with faceless man in headphones, a black hoodie and joggers and using a futuristic computer with a view of a futuristic neon-lit city on a rainy night
Macro Shot
/imagine macro photography, snowflake, super zoomed in, photorealistic, 8k –q 2
Minimalism Art
/imagine tarot card illustration, gemini as the lovers, minimalistic style, elegant golden lines, black card
Mood Boards
/imagine fashion moodboard for winter tone clothes inspired by the outdoor travel
Paper Style Art
/imagine paper cut craft, 3d paper illustration, realistic, origami winter background, mount everest, colorful, very detailed, 8k, full size, +distance –v 4 (Version 1)
/imagine paper cut craft, 3d paper illustration, realistic, origami winter background, mount everest, colorful, very detailed, 8k, full size, +distance –v 4 (Version 2)
Pixel Art
/imagine 32 byte, japanese car , busy city, super high resolution , psychedelic colors , pixel art , retro cartoon (Version 1)
/imagine 32 byte, japanese car , busy city, super high resolution , psychedelic colors , pixel art , retro cartoon (Version 2)
/imagine 16-bit pixel art, downtown, many cars, cinematic still, hdr –ar 3:2 –q 2
Portraiture Art
/imagine A beautiful urban fantasy heroine, focus face and eyes, flawless face, blonde hair, pink leather trench coat, photography, portrait, f 1.2, –ar 2:3 –no hands
/imagine an Aztec princess in futuristic design, character portrait, detailed, realistic, photography, brown hair, surprised expression, fantasy, glass painting, beautiful, soft evening lighting –ar 2:3
/imagine portrait photography, female raiden genshin impact, half body, looking at camera, sword on back, rules of thirds, f 1.2, focus on eyes, blur background, busy cherry blossom background, light purple lighting, hyperrealism –ar 2:3 –no hands
/imagine a full body portrait, blonde attractive woman, in the style of artgerm lau, full body, red dress, ultra realistic
Product Design
/imagine futuristic tennis shoe concept, orange and blue, liquid form, commericial photo, 8k
/imagine Midway pinball, intrincated retro pinball, glowing innerlight intricate detail, platforms, mazes and lights / glowing innerlight intricate details / fine shiny metallic edges / photorealistic / cinematic lighting –v 4 –q 2
/imagine 4 pillows for sofa, brown white, arrow and geometric shapes arrays, 8k, product photo
/imagine Design futuristic example for trading card game cards, intricate, dark background (Version 1)
/imagine Design futuristic example for trading card game cards, intricate, dark background (Version 2)
Stained Glass Art
/imagine stained glass shiba, cinematic, very realistic, lots of detail, blue tones, cinematic lighting
/imagine flaming music clef, hexagonal, stained glass window, epic background, In the style of studio ghibli, insane high detailed beautiful composite, sharp line work, clear focus, dramatic lighting, epic composition, sharp contours, graphic detail, colored, geometric filigree in style of Alphonse Mucha, David zapora –v 4 (Version 1)
/imagine flaming music clef, hexagonal, stained glass window, epic background, In the style of studio ghibli, insane high detailed beautiful composite, sharp line work, clear focus, dramatic lighting, epic composition, sharp contours, graphic detail, colored, geometric filigree in style of Alphonse Mucha, David zapora –v 4 (Version 2)
/imagine luminous stained glass, zelda, Solid background –ar 3:2 (Version 1)
/imagine luminous stained glass, zelda, Solid background –ar 3:2 (Version 2)
Watercolor Style
/imagine paint splash and stains on white paper that form a colorful outdoor scene, set in the mount everest–ar 2:3
Web Design Elements
/imagine logo icon for cooking app
/imagine orange tabby kitten as a squared app logo design, flat vector app
/imagine illustration for a tech company, baker at a bakery, style of behance, vector illustration –q 2
/imagine modern app, UI/UX, UI, UX, beautiful design, orange and blue, neumorphic style, AV control, automation, SAVI Controls
/imagine book website design, ui/ux, book collection website, homepage (Version 1)
/imagine book website design, ui/ux, book collection website, homepage (Version 2)
Wildlife Art
/imagine a bird suspended in the air, ripples and splash on surface, sunlight gleaming, with sparkling crisp radiant reflections, Canon 35mm lens, hyperrealistic photography, style of unsplash –q 2
Breakdown of /imagine Command and Text Prompt for Midjourney
Now you understand the basic capabilities of Midjourney, YES! those were basic examples of what this AI art generator can do.
Let’s breakdown the structure of a text prompt for Midjourney.
What is a prompt?
A prompt is a request or suggestion to do something or provide information.
What is a text to image prompt?
A text to image prompt within the context of an AI art generator is a request for the AI system to generate an image based on a given text input.
The text input could be a word, phrase, or a more complex description of the desired image, and the AI system will use this input to generate an original image that aligns with the given prompt.
This type of prompt allows users to specify the content or theme of the generated image and can be used to create a wide range of visual styles and subjects.
Basic Structure of a command within Midjourney
Here’s an image example of the /imagine command from Midjourney Gitbook Quick Start Guide.
Midjourney’s ai art generator command it’s currently broken down into four sections:
[/imagine Command] – [Image prompt] – [Text Prompt] – [Parameters]
For the rest of the blog we’ll focus on 3 of the 4 sections:
[/imagine Command]
[Text Prompt]
[Parameters]
We’ll save the [Image prompt] for another time.
Section 1 – [/imagine Command]
When typing /imagine in the Discord chat the Midjourney Bot will reveal the prompt command for you to enter the [Image prompt] and/or [Text Prompt] and/or with/without [Parameters]
With these 3 sections, your possibilities in generating ai generated art are literally endless.
Section 2 – [Image prompt]
Since this is an in-depth guide for text prompts we’ll reserve this topic for another time.
However, don’t forget that simply uploading an image and providing the URL to the image as a reference will also create beautiful results of AI-generated art.
Section 3 –[Text Prompt]
The text prompt section in Midjourney is the meat and potatoes of how to create an image or artwork by typing in a descriptive word or words.
The more descriptive the better the outcome.
Commas are used to divide each descriptive word, thus enabling the creation of AI artwork.
For example, you might type in “/imagine prompt:japanese car , busy city, super high resolution , psychedelic colors , pixel art , retro cartoon”
Midjourney will generate into an ai art for you to see.
Although you are not required to divide your descriptive words with commas, doing so allows Midjourney bot to understand that each word is its own distinct thought.
For example, Type in “/imagine prompt:japanese car busy city super high resolution psychedelic colors pixel art retro cartoon”
Even without commas results comes out amazing.
You can also use something called “text weights” to give some words or phrases more important than others.
Instead of commas (,) you can replace them with two colons (::) and numeric values after descriptive words, to let Midjourney know what to focus on.
For example, you might type in “/imagine prompt:japanese car::5 busy city::1 super high resolution::4 psychedelic colors::1 pixel art::1 retro cartoon::1″
It lets Midjourney focus in on it’s creation
The “japanese car” and “super high resolution” words should have more weight and priority for Midjourney.
Section 4 – [Parameters]
Parameters can be thought of as “inputs” for the command.
While some are necessary, like a prompt, others are optional and will modify how the prompt is interpreted and the image is created.
These parameters may also go by many other names such as “switches” or “flags”.
Both [image prompt] and [text prompt] are parameters.
For example, a full image command could feature multiple components, for instance an image URL, specific text and weights plus other parameters:
/imagine prompt: street racing car, in busy city, vibrant –v 4a –hd
You may customize your command’s output by finishing it off with any of the parameters listed below, each designated by the “–” delimiter.
To guarantee the accuracy of your output, make sure to include all parameters at the end of each command. Neglecting this essential step may result in a product that is unusable or has severe glitches.
Quick Starter Guide on Midjourney Workflow
Here’s a quick reminder of the step-by-step guide in Midjourney workflow.
1. Now you are pump to start making AI art!
Remember that only a certain number of prompt options will be available during your trial period.
Around 25 free images can be created before you need to purchase a full membership in order for continuous use.
To save money and create the best outcome possible, take some time beforehand and think about what is desired on Midjourney.
If there’s any doubt or confusion along the way, don’t hesitate – type “/help” into search bar for helpful tips and tricks.
2. When you’re ready, type ‘/imagine’ in the Discord chat for your newbies room.
This will prompt a field where you can provide an image description; so be as descriptive and precise as possible!
If there’s a specific aesthetic or style that you’d like to capture with this image, include it in the description – the AI is best able to produce good results when given adequate specificity.
Refer back to the command and prompt section if you need some guidance on text prompting.
Remember, there are terms of conduct to follow here – but if you keep it clean, everything should flow smoothly.
When finished crafting your request, hit Enter and watch your vision come alive!
3. After you give Midjourney a minute to generate your images, the AI will provide four unique versions of your description.
You’ll then see U and V buttons labeled 1 through 4 located beneath these pictures; each number corresponds with one of the generated images.
Now that everything is set up, you have multiple ways to continue forward.
Choosing U will upscale that particular image into a larger, more defined version.
U1 – Upscale TOP LEFT Image
U2 – Upscale TOP RIGHT Image
U3 – Upscale BOTTOM LEFT Image
U4 – Upscale BOTTOM RIGHT Image
Choosing V will create an all-new image based on the present image that you choose.
V1 – Generates 4 New Variations of the TOP LEFT Image
V2 – Generates 4 New Variations of the TOP RIGHT Image
V3 – Generates 4 New Variations of the BOTTOM LEFT Image
V4 – Generates 4 New Variations of the BOTTOM RIGHT Image
You will also see a refresh button to the side to request a new set of images.
Refresh Button – Regenerates all 4 images.
Keep in mind that each of these choices will use up some of your available free prompts, so only do it if you are sure you want to proceed.
4. If you find a picture that captivates your imagination, then select the ‘U’ option to upscale it and uncover new options.
You can make modifications on the image, or even advance it up to its maximum capacity before downloading if desired.
If you’re not completely satisfied with what you have, go ahead and choose a light upscale redo of it too!
When finished, just send it to yourself by clicking the envelope emoji which will allow for both savings as well as accessing direct downloads whenever required in the future.
Optional Step: If you plan on using Midjourney frequently, the easiest way to stay up-to-date is by joining its Discord channel and typing in ?/subscribe.?
Simplified Image Size within Midjourney
The next question many Midjourney users and onlookers want to know is how large or if these generated images can be printed.
The answer is yes and no. Hear me out…
The default size of the initial 2×2 grid is 512 x 512 pixels.
To illustrate the significance of this, consider that printing out at 300 DPI (dots per inch) with a resolution of 512×512 is comparable to 1.71 x 1.71 inches in print form.
There is also the argument of printing out the 512 x 512 pixels with 72 DPI (equivalent to 7.1 x 7.1 inches) or 96 DPI (equivalent to 5.3 x 5.3 inches), but printing with those metrics the image will be blurry and not very crisp.
For the rest of the sizes we’ll calculate with 300 DPI.
The default upscale size is 1024 x 1024 pixels.
[Printing at 300 DPI with 1024 x 1024 pixels equivalent to 3.4 x 3.4 inches]
The default upscale to max size is 1664 x 1664 pixels.
[Printing at 300 DPI with 1664 x 1664 pixels equivalent to 5.5 x 5.5 inches]
In a recent Midjourney Office Hours, David Holz – the founder of Midjourney did express optimism about furthering the upscale within Midjourney AI.
Key Parameters and Terms to Use Midjourney Effectively
Let’s dive deeper into Midjourney AI and its capabilities.
While we merely touched upon the parameters in the previous section, they will become increasingly impactful as newer versions of Midjourney are released. Parameters will have a greater influence on how AI art is generated within the AI art generator.
Don’t forget to add the parameter at the end of your command to use it.
Here is an extensive list of parameters, with explanations on how to use them for command to generate ai art.
Size: Width and Height
–aspect or –ar This feature allows you to ensure that the generated image has an aspect ratio as close as possible to your desired one; it can be a reduced ratio, such as 3:2, or even reducible ratios like 1920:1280.
If no specific value is indicated, then images will have a square aspect ratio –ar 1:1.
To ensure that your image has a 3:2 aspect ratio, simply add –ar 3:2 to the end of your command line prompt.
Example of –ar 3:2
To ensure that your image has a 2:3 aspect ratio, simply add –ar 2:3 to the end of your command line prompt.
Example of –ar 2:3
Please be aware that the ratio will adjust to an even measure., for example, –ar 3:2 is the same as –ar 1920:1280.
Regardless of the situation, we can achieve a resolution of 384 x 256 for the initial generation, followed by 1536 x 1024 after the first upscale, and then 1920 x 1280 at maximum upscale.
If you’re looking for the best resolution, –ar is your go-to.
These aspect ratios can only be used with v3 and older. Not currently compatible with v4.
–ar 9:16 Smartphone (vertical)
–ar 16:10 Standard desktop screen
–ar 21:9 or –ar 24:10 Widescreen
–ar 10:14 A paper standard (A0-A10)
–ar 16:9 HD film / smartphone (horizontal)
–ar 3:2 Classic film
Algorithm Modifers
Version 1
–v 1 is the original Midjourney Bot Algorithm.
You are able to achieve a more abstract design that may be better suited for macro images or texture-oriented projects with this model.
Version 2
–v 2 uses the original Midjourney Bot Algorithm as well but with some minor improvements.
Version 3
–v 3 this is when AI art generators started to become more widely recognized.
Version 4
–v 4 is the default model unless changed in the setting mode.
Version 4 has two options, –v 4a or –v 4b
This version model was trained with AI supercluster.
Version Niji
–niji is the newest model and for the anime-style art generation. This is my personal favorite model.
High Definition
–hd For larger images, this algorithm provides a viable solution that can produce even superior results.
If you’re looking for better upscaling resolutions, it’s highly recommended to opt for a multiple of 64 (or 128 if using –hd) as this prevents undesirable results.
Furthermore, it is not compatible with–hd, –stylize, and –q 5.
Stylize Values
–stylize or –s Fine-tune your art generation to match the creative vision you have in mind by selecting any integer between 625 and 60000.
Here are a few samples of what this looks like:
–stylize 625 If you would like to take a break from the creative process, then turning it off may be an ideal solution.
–stylize 1250 If you’re looking for something that has flexible yet pleasing styling, this is the perfect choice for experienced users.
–stylize 2500 The default value, so you don’t have to specify this.
–stylize 20000 If you’re looking for a subtle, yet dynamic twist to your text, let it ‘take over’ and see what image gets generated.
–stylize 60000 Step away from the steering wheel and see what surprises arise. You never know where this journey could take you – it could be better than anything you ever imagined!
Quality Values
–quality or –q reduces the amount of time spent creating your image.
As a reminder, only use one of the provided values. Any other input will be automatically rounded to the nearest valid value.
–quality 0.25 If you’re looking to save some credits while still utilizing Midjourney, try this quality option. You’ll get 4x faster results at the cheapest rates.
–quality 0.5 Less detailed results but 2x faster / cheaper.
–quality 1 The default value, you do not need to add at the end.
–quality 2 Currently, the most used parameter among Midjourney AI users is one that offers more comprehensive results; however, it does come at double the cost and runs twice as slow. (2 GPU minutes per /imagine).
–quality 5 If you genuinely want to challenge the capabilities of artificial intelligence, use this quality value. It might encourage creative and comprehensive results… however, it could also yield worse outcomes! (5 GPU minutes per /imagine).
–stop To stop performance, set the generation to a lower percentage in the range of 10-100; note that this does not work for upscales.
-no a new prompt has been added that allows for further customization. For instance, using the “–no grass” parameter attempts to remove any grass from your photo.
Negative Prompting
–no a new prompt has been added that allows for further customization. For instance, using the “–no grass” prompt attempts to remove any grass from your photo.
Chaos Prompting
–chaos <number> Another recently newly added prompt. The higher the value placed between 0-100, the more creative and unpredictable results you will receive within the grid.
General Modifiers
–beta utilizes a new experimental algorithm. This model requires longer processing time, but produces two outputs that are optimized for non-square resolutions.
–seed Establishing a seed has the potential to yield more consistent and reproducible output when needing to generate a prompt again.
Seeds must be within the range of any number between 0 and 4294967295, any positive integer number will do. If you don’t specify one, an unpredictable random seed will automatically be chosen in its place behind the scenes.
–sameseed This prompt ensures that the seed impacts all images across the grid to be more identical. But, if left unused, each image in the grid will employ a distinct “slice” from the seed resulting in even higher variation. Currently cannot use in Midjourney v4.
–iw With this feature, you can set the image prompting weight in relation to the text weight. Default value is –iw 0.25
–video Get a progress video sent directly to you with an ??-triggered DM (simply react with the envelope to obtain the link). Please note that this does not work for upscales.
–uplight Uses the light upscaler when selecting the U buttons. With this technique, results appear closer to the original image and fewer details are added during upscaling. It is especially useful for faces and surfaces with a smooth texture.
You probably noticed by now there’s no way to add, subtract or change words within a command or prompt during the variation process, right? Wrong!
There the Remix feature allows you to edit your prompt during the variation process, and toggle it on or off using /prefer remix (or toggled using /settings).
When switched on, Remix provides the ability to modify prompts when creating variants of images. You can adjust the entire prompt or simply make minor changes for a unique approach, then continue on your variation process.
Now you see why parameters will become even more of a significant role within Midjourney?
They’re basically like the buttons on a gaming controller to get the AI to generate more accurate results.
Tips for Text Prompt for Midjourney
- Use objects that are easily recognizable and have many images available online.
Example: “A photo-realistic image of the Eiffel Tower” - Incorporate strong emotions or mystical themes in your prompts.
Example: “An image of a mysterious forest that evokes a sense of wonder” - Clearly describe the style you are looking for.
Example: “A vintage-inspired poster of a 1950s diner” - Use the names of specific artists to influence the style of the output.
Example: “A painting in the style of Salvador Dali” - Combine the names of different artists to create unique styles.
Example: “A surreal landscape in the style of Salvador Dali and Rene Magritte” - Use positive language and avoid negatives.
Example: “A futuristic car” instead of “A car that is not from the past” - Be clear and specific about what you want.
Example: “A detailed illustration of a dragon with purple scales and yellow eyes” - Specify the composition if you have a specific one in mind.
Example: “A close-up shot of a tiger’s face” - Avoid including too many small details in your prompts as it can overwhelm the model.
Example: “A chaotic scene of a monkey riding a unicycle and juggling knives on a tightrope”
instead of “A monkey riding a unicycle, juggling knives, on a tightrope, in a circus, in the middle of a storm, with a background of a city” - Combine two distinct concepts in new and unique ways.
Example: “A cyberpunk wizard” - Use specific nouns or numbers.
Example: “Three cyberpunk wizards” instead of “cyberpunk wizards” - Avoid prompts that require the model to make significant assumptions or inferences.
Example: “A solarpunk city filled with holograms”
instead of “An optimistic vision of an augmented reality future”
Conclusion
Now that you know how to use Midjourney’s text-to-image prompts, it’s time for you to start creating your own AI art!
Remember to stay creative and experiment with a variety of prompts.
With a little bit of practice, you’ll be able to create amazing works of art using this revolutionary AI tool. Good luck!
Happy Generative Art making!
Resources I Use
Resource 1: Midjourney Documentation Gitbook. ? Read More
Resource 2: Midjourney Bot in Discord. ? Learn More
Resource 3: Discord Server. ? Learn More
