Show Menu
Cheatography

The Image Framework - A CustomGPTs Guide Cheat Sheet (DRAFT) by

Swiss army knife for DALL-E. COMMANDS: /help - /mode - /style - /images - /ratio - /dimension - /showprompt - /variation - /emotion - /exclude - /color - /txt - /background or /bg - /foreground or /fg - /nobg

This is a draft cheat sheet. It is a work in progress and is not finished yet.

/style [preferred style]

/style illust­ration
Creates an illust­ration
/style hand-drawn
Creates a hand-drawn image
/style cinematic
Creates a cinematic image
/style floral
Creates a floral image
/style vibrant
Creates a vibrant image
You can also use your own style or one of those predefined styles:

fractal, realistic, minima­list, abstract, waterc­olor, fibonacci, steampunk, cartoon, glossy, neon, glitch, vintage, futuri­stic, mystical, ethereal, photor­eal­istic, digital, pixel art, HDR, manga, botanical, nautical, tropical, urban, landscape, nature, underw­ater, histor­ical, fantasy, sci-fi, horror, romantic, satirical, magical, cityscape, countr­yside, space, dystopian, utopian, post-a­poc­aly­ptic, renais­sance, modern

/mode [preferred mode]

/mode sticker
Creates an image with a slight emphasis on the upper body of the subject. This mode is designed for images that need to be recogn­izable even when scaled down to smaller sizes, like stickers. It ensures that key features remain visible and distinct.
/mode emote
Focuses signif­icantly on the head or facial expres­sions of the subject. This mode is ideal for images that will be drasti­cally reduced in size, such as digital emotes or emojis, where facial expres­sions or head details need to be clear and expres­sive.
/mode portrait
Creates a detailed, close-up image focused on the face or upper body of a subject, ideal for profile pictures or avatars.
/mode icon
Generates a simpli­fied, iconic repres­ent­ation of the subject, emphas­izing clear, recogn­izable features suitable for small icons or avatars.
/mode banner
Generates an image with dimensions and focus suitable for headers or banners, with the main subject centered and space for text or additional elements.
/mode collage
Produces a composite image made from multiple photog­raphs or elements, arranged artist­ically, perfect for storyt­elling or thematic presen­tat­ions.
You can also describe and use your own mode with natural language, and The Image Framework will take it into account.

/text [on] or [off]

/text on
Creates an image and forces text to appear
/text off
Creates an image without any text at all (exper­ime­ntal)
"­/text off" is still experi­­me­ntal. Sometimes it causes the exact opposite, but for the most part it is reliable. It is best to use it only when there is no other way.
 

/images [count]

/images 1 (default)
Creates one image
/images 5
Creates five images
/images 10
Creates ten images
10 is the maximum. DALLE has separate rate limits:

200 is the DAILY maximum, there is a 15 min and a 60 min rate limit to help with load. This limit is variable and depends on the load, so during off hours it is higher.

/variation [0 to 100]

/variation 0
Creates images with little to no variation between them
/variation 73
Creates images with lots of variation between them
/variation 100
Creates images with so much variation that every image is different from another
This command can only be used together with the "­/im­age­s" command. The higher the number, the more variation there is between generated images.

/color [color]

/color green
Creates an image, mainly using the color green
/color red and blue
Creates an image, mainly using the colors red and blue
/color sunflowers
Create an image, mainly using the colors of a sunflower (brown and yellow)
/color HEX #41FDBC
Creates an image, mainly using the given color hex value
The image mostly uses the given color or colors. The colors can be given in plain text, RGB values or HEX values.

/tint [tint]

/tint green
Creates an image with a green tint
/tint red and blue
Creates an image with a red and blue tint
/tint sunflowers
Create an image, tinted in the colors of a sunflower (brown and yellow)
/tint HEX #41FDBC
Creates an image, tinted in the given color hex value
The image gets a tint in the given color or colors. The colors can be given in plain text, RGB values or HEX values. You can add phrases like "­strong tint" or "­sli­ghtly tinted­" to the prompt to adjust the strength.

/nobg

/nobg
Generates an image with a one-color background
Great for making the image transp­arent by keying out the background
 

/exclude [anything]

/exclude animals
Excludes all sorts of animals
/exclude nature, trees, bushes and forests
Excludes nature elements, especially trees, bushes and forests
What is specified with the exclude command will be ignored during image genera­tion. You can use natural language to describe what should be excluded.

This feature is still experi­mental. Sometimes it causes the exact opposite, but for the most part it is reliable. It is best to use it only when there is no other way.

/dimension [2D] or [3D] or [4D] or [5D]

/dimension 2d
Creates a 2D image
/dimension 3d
Creates a 3D image
/dimension 4d
Creates a 4D image (exper­ime­ntal)
/dimension 5d
Creates a 5D image (exper­ime­ntal)

/ratio [aspect ratio]

/ratio square (default)
Creates an image with height and length being the same
/ratio 16:9
Creates a 16:9 image
/ratio 4:3
Creates a 4:3 image
/ratio Tube-TV
Creates an image with the dimensions of old Tube-TV's
Use your preferred aspect ratio or use natural language to describe which dimensions the generated image should have.

/showp­rompt

/showp­rompt
If used, the prompt that The Image Framework created and sent to DALLE is shown. This can then be copied, pasted, and slightly adjusted if you want similar results. Works great with Midjourney too.

/foreg­round [element] or /fg [element]

/fg kangaroo
Creates an image with a kangaroo in the foreground
/fg donut on a table
Creates an image with a donut on a table in the foreground

/backg­round [element] or /bg [element]

/bg antarctica
Creates a picture with an Antarctic scene in the background
/bg storefront in a big city
Creates an image with a storefront in a big city in the background

/emotion [emotion]

/emotion happy
The generated image could display a sunny day or people smiling
/emotion sad
The generated image could display a rainy day or someone looking sad
Hint: You can use any emotion with natural language.