Also written: infographic, info graphic, explainer graphic
A graphic that explains information with icons, sections and short labels.
What the AI does
Builds a clean layout with icons, numbered sections and simple charts.
When to use it
For explaining steps, facts or comparisons at a glance.
Where it works
Not an official command: the model simply reads '/infographic' as the words 'infographic'. Typing '/infographic' gives the same result as writing 'infographic' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Trusting the numbers or text; always check and fix them.
Putting too many facts in one image.
Example
An infographic about the water cycle, clean infographic layout, simple icons, numbered sections, flat vector style, clear hierarchy
/beforeafter
Also written: before and after, before/after, split comparison
Two versions side by side showing a change.
What the AI does
Splits the image into a 'before' half and an 'after' half of the same subject.
When to use it
For makeovers, renovations, cleaning, editing and transformations.
Where it works
Not an official command: the model simply reads '/beforeafter' as the words 'before and after comparison'. Typing '/beforeafter' gives the same result as writing 'before and after comparison' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Not describing both states clearly.
Using it for real results claims; AI images are illustrations, not proof.
Example
A messy garage on the left and the same garage tidy on the right, before and after split comparison, same subject shown in two halves, clear dividing line
/poster
Also written: poster, movie poster, event poster
A bold, eye-catching vertical design with a title area.
What the AI does
Creates a strong central image, dramatic composition and space for a title.
When to use it
For events, movies, campaigns and wall art.
Where it works
Not an official command: the model simply reads '/poster' as the words 'poster design'. Typing '/poster' gives the same result as writing 'poster design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Asking for long text on the poster.
Not saying the theme or mood.
Example
A poster for a summer jazz festival, poster design, bold central image, strong composition, space for a title at the top
/logo
Also written: logo, logo design, brand mark
A simple symbol or wordmark for a brand.
What the AI does
Makes a clean, simple mark on a plain background, usually flat and centered.
When to use it
For brainstorming brand ideas and early concepts.
Where it works
Not an official command: the model simply reads '/logo' as the words 'logo design'. Typing '/logo' gives the same result as writing 'logo design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. AI Compass also has a Logo Designer mode (/logo) in chat. In AI Compass chat, /logo is a real command that opens the Logo Designer mode.
Common mistakes
Expecting an editable vector file; results are pictures.
Asking for long or complex brand names; letters may be wrong.
Example
A logo for a small bakery called Sunrise, simple flat logo design, clean vector-style mark, centered on a plain white background
/mockup
Also written: mockup, mock-up, product mockup
A realistic preview of a design placed on a real-looking product or screen.
What the AI does
Shows your design on an object like a phone, mug, bag or poster frame.
When to use it
For presenting designs, apps or merchandise ideas.
Where it works
Not an official command: the model simply reads '/mockup' as the words 'product mockup'. Typing '/mockup' gives the same result as writing 'product mockup' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Expecting your exact logo or screenshot to appear; describe it or use image editing.
Forgetting to name the object.
Example
A mockup of a coffee mug with a mountain illustration, realistic product mockup, design shown on the object, clean studio background, soft shadows
/thumbnail
Also written: thumbnail, YouTube thumbnail, video thumbnail
A bold preview image for a video.
What the AI does
Creates a high-contrast, simple image with one clear subject and space for a short title.
When to use it
For video covers and social previews.
Where it works
Not an official command: the model simply reads '/thumbnail' as the words 'video thumbnail'. Typing '/thumbnail' gives the same result as writing 'video thumbnail' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Adding long text; keep it to two or three words and add it in an editor.
Too many small details that vanish at small size.
Example
A thumbnail for a cooking video about homemade pizza, eye-catching video thumbnail, one bold subject, high contrast, bright colors, space for a short title
/carousel
Also written: carousel, multi-slide post, slides
A set of images meant to be swiped through in order.
What the AI does
Keeps a consistent style across slides; in most tools you prompt each slide separately.
When to use it
For step-by-step social posts and stories.
Where it works
Not an official command: the model simply reads '/carousel' as the words 'carousel slide series'. Typing '/carousel' gives the same result as writing 'carousel slide series' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. AI Compass's Carousel tab makes a set of 2-10 images on one topic.
Common mistakes
Expecting one prompt to produce a full matching set in every tool.
Changing the style between slides.
Example
Slide one of a carousel about morning routines, carousel slide, consistent clean design, same colors and style across the series
/banner
Also written: banner, header image, cover image
A wide image used at the top of a website, channel or profile.
What the AI does
Creates a wide layout with the subject to one side and empty space for text.
When to use it
For website heroes, channel covers and profile headers.
Where it works
Not an official command: the model simply reads '/banner' as the words 'banner design'. Typing '/banner' gives the same result as writing 'banner design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. AI Compass also has a Banner mode (/banner) in chat. In AI Compass chat, /banner is a real command that opens the Banner mode.
Common mistakes
Forgetting it needs a wide shape (choose landscape where available).
Adding long text inside the image.
Example
A banner for a travel blog with mountains and a lake, wide banner layout, subject on one side, clean empty space for text, balanced design
/storyboard
Also written: storyboard, comic panels, shot sequence
A set of panels sketching the shots of a story or video.
What the AI does
Draws several numbered panels showing scenes in order.
When to use it
For planning videos, ads and comics.
Where it works
Not an official command: the model simply reads '/storyboard' as the words 'storyboard'. Typing '/storyboard' gives the same result as writing 'storyboard' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Expecting perfect continuity between panels.
Describing too many scenes for one image.
Example
A storyboard of a cat sneaking into a kitchen, storyboard sheet, six sketch panels in a grid, simple shots in order, pencil style
/icon
Also written: icon, app icon, icon design
A small, simple symbol for an app or interface.
What the AI does
Makes a simple, centered shape with bold color that reads at small size.
When to use it
For app icons, buttons and simple symbols.
Where it works
Not an official command: the model simply reads '/icon' as the words 'app icon'. Typing '/icon' gives the same result as writing 'app icon' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Adding fine detail that disappears at small size.
Putting words inside the icon.
Example
An app icon for a weather app, simple app icon, bold centered symbol, rounded square, flat colors, clean shapes
/pattern
Also written: pattern, seamless pattern, tile
A repeating design that can tile without visible edges.
What the AI does
Repeats small motifs evenly across the image.
When to use it
For fabric, wrapping paper, backgrounds and wallpapers.
Where it works
Not an official command: the model simply reads '/pattern' as the words 'seamless pattern'. Typing '/pattern' gives the same result as writing 'seamless pattern' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Expecting a perfectly seamless tile from every tool; check the edges.
Using one large subject instead of small repeated motifs.
Example
A pattern of lemons and leaves, seamless repeating pattern, small evenly spaced motifs, flat colors
/wallpaper
Also written: wallpaper, background, lock screen
A full-screen background image for a phone or computer.
What the AI does
Creates a clean, attractive full-screen image without clutter in the middle.
When to use it
For phone and desktop backgrounds.
Where it works
Not an official command: the model simply reads '/wallpaper' as the words 'phone or desktop wallpaper'. Typing '/wallpaper' gives the same result as writing 'phone or desktop wallpaper' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Forgetting the shape: portrait for phones, landscape for desktops.
Busy details where icons will sit.
Example
A calm mountain lake at dawn as a phone wallpaper, beautiful wallpaper, full-screen composition, calm uncluttered center, rich colors
/quotecard
Also written: quote card, quote image, quote post
A simple background image designed to hold a short quote.
What the AI does
Creates a calm background with a clear area for text.
When to use it
For social posts with a quote or tip.
Where it works
Not an official command: the model simply reads '/quotecard' as the words 'quote card'. Typing '/quotecard' gives the same result as writing 'quote card' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Letting the AI write the quote text into the image; add it in an editor.
Using a busy background.
Example
A quote card background with soft morning light, quote card background, calm soft gradient, clear empty center for text, minimal design
/flyer
Also written: flyer, leaflet, promo flyer
A one-page promotional design for an event or offer.
What the AI does
Lays out an eye-catching image with sections for title and details.
When to use it
For events, shops, classes and announcements.
Where it works
Not an official command: the model simply reads '/flyer' as the words 'flyer design'. Typing '/flyer' gives the same result as writing 'flyer design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Relying on the AI for dates, prices and contact details; add them yourself.
Too much text.
Example
A flyer for a neighborhood yoga class in the park, flyer design, eye-catching main image, clear title area, space for details below
/collage
Also written: collage, photo collage, mixed media collage
Several images or cut-outs combined into one picture.
What the AI does
Arranges multiple images or paper cut-outs in one layout.
When to use it
For mood boards, memories and creative posters.
Where it works
Not an official command: the model simply reads '/collage' as the words 'collage'. Typing '/collage' gives the same result as writing 'collage' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Expecting your own photos to appear; describe them instead.
Too many items, so nothing stands out.
Example
A travel collage of a summer trip to the sea, collage of several photos and paper cut-outs, layered mixed media layout
/comparison
Also written: comparison, side by side, versus
Two or more things shown next to each other to compare.
What the AI does
Places items in equal panels so differences are easy to see.
When to use it
For product choices, styles and 'this vs that' posts.
Where it works
Not an official command: the model simply reads '/comparison' as the words 'side-by-side comparison'. Typing '/comparison' gives the same result as writing 'side-by-side comparison' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Trusting any labels or numbers the AI adds.
Not keeping the angle and lighting the same.
Example
Two coffee drinks compared side by side: latte and cappuccino, side-by-side comparison, equal panels, same angle and lighting, clean layout
/diagram
Also written: diagram, schematic, flowchart
A simple drawing that explains how something works.
What the AI does
Creates labeled shapes, arrows and parts in a clean style.
When to use it
For explaining processes, parts and ideas.
Where it works
Not an official command: the model simply reads '/diagram' as the words 'diagram'. Typing '/diagram' gives the same result as writing 'diagram' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Relying on labels and arrows being correct; check them.
Asking for a very complex technical drawing.
Example
A diagram of how a plant makes food from sunlight, simple clean diagram, labeled parts, arrows showing flow, flat minimal style
/timeline
Also written: timeline, history timeline, roadmap graphic
A graphic that shows events in order along a line.
What the AI does
Draws a line or path with milestones and simple icons.
When to use it
For history, project steps and personal journeys.
Where it works
Not an official command: the model simply reads '/timeline' as the words 'timeline graphic'. Typing '/timeline' gives the same result as writing 'timeline graphic' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Expecting correct dates or facts; add and check them yourself.
Too many milestones.
Example
A timeline of the steps to plant a garden, horizontal timeline graphic, milestones with simple icons, clean flat design
/albumcover
Also written: album cover, cover art, music cover
Square artwork for a music release.
What the AI does
Creates a bold square image with a strong mood that fits the music.
When to use it
For singles, albums, playlists and podcasts.
Where it works
Not an official command: the model simply reads '/albumcover' as the words 'album cover art'. Typing '/albumcover' gives the same result as writing 'album cover art' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Copying a famous artist's cover style or name.
Adding long text.
Example
An album cover for a calm piano album, square album cover art, bold striking image, strong mood, room for a title
/bookcover
Also written: book cover, cover design, novel cover
A vertical cover image for a book or ebook.
What the AI does
Creates a striking image that hints at the story, with space for title and author.
When to use it
For ebooks, story drafts and cover concepts.
Where it works
Not an official command: the model simply reads '/bookcover' as the words 'book cover design'. Typing '/bookcover' gives the same result as writing 'book cover design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. Text inside images is often misspelled; keep words short or add them later in an editor.
Common mistakes
Letting the AI write the title text; add it in an editor.
Not giving the genre.
Example
A book cover for a mystery novel set in an old lighthouse, book cover design, vertical layout, striking central image, space for title and author name
/packaging
Also written: packaging, package design, box design
The design on a box, bag, bottle or label.
What the AI does
Shows a product package with a designed label in a realistic setting.
When to use it
For product concepts and label ideas.
Where it works
Not an official command: the model simply reads '/packaging' as the words 'packaging design'. Typing '/packaging' gives the same result as writing 'packaging design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Expecting exact brand text or legal labels.
Using real brand names or logos.
Example
Packaging for organic honey in a glass jar, product packaging design, clean modern label, realistic box on a studio background
/tshirtdesign
Also written: t-shirt design, shirt print, merch design
A graphic made to be printed on clothing.
What the AI does
Creates a centered graphic, often on a plain background, sometimes shown on a shirt.
When to use it
For merch ideas and print concepts.
Where it works
Not an official command: the model simply reads '/tshirtdesign' as the words 't-shirt print design'. Typing '/tshirtdesign' gives the same result as writing 't-shirt print design' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Using copyrighted characters or brand logos.
Too much fine detail for printing.
Example
A t-shirt design of a mountain and a moon, t-shirt print design, bold centered graphic, limited colors, plain background
/emoji
Also written: emoji, emoji style, emoticon
A tiny, round, expressive cartoon in the style of emoji.
What the AI does
Draws a simple, glossy, very expressive character or object.
When to use it
For reactions, stickers and fun icons.
Where it works
Not an official command: the model simply reads '/emoji' as the words 'emoji style'. Typing '/emoji' gives the same result as writing 'emoji style' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator.
Common mistakes
Copying a specific company's emoji set.
Adding scenery around it.
Example
A laughing cactus emoji, emoji style, simple glossy cartoon, big expression, plain white background
/avatar
Also written: avatar, profile picture, pfp
A picture of a person or character for a profile.
What the AI does
Creates a centered head-and-shoulders portrait with a simple background.
When to use it
For profile pictures and gaming or forum avatars.
Where it works
Not an official command: the model simply reads '/avatar' as the words 'profile avatar'. Typing '/avatar' gives the same result as writing 'profile avatar' in plain words. Works in GPT image, Gemini/Imagen, Midjourney, Stable Diffusion and AI Compass's generator. AI Compass also has an Avatar mode (/avatar) in chat with style, mood and background choices. In AI Compass chat, /avatar is a real command that opens the Avatar mode.
Common mistakes
Asking for a real person's face or a celebrity.
Busy backgrounds that look messy when small.
Example
An avatar of a smiling astronaut, profile avatar, centered head and shoulders, friendly expression, simple soft background
Every guide and dictionary card is also in the AI Compass app, with saved items, search and “Try it” buttons that open chat, image or video with the example filled in. Reading is free; creating uses credits, and the price is shown before you start.