Conjure Suite
Write what you want to see. Or... provide a reference image then write what you want to see for greater creative control.
Last updated
Write what you want to see. Or... provide a reference image then write what you want to see for greater creative control.
Last updated
The Conjure suite is dedicated to turning words into images. Simply describe the image you want to see and hit conjure to see the magic happen.
Before hitting conjure to generate your image, you can change the output aspect ratios, and if you're uncertain, upon hovering you can preview the shape of the aspect ratio before selecting to ensure the proper selection has been made:
Note that 4:5 and 9:16 are the most popular for Instagram.
This suite offers a robust variety of predetermination tools (doodler, x-ray, style guide, skeleton, ect...) Though each tool varies in its use cases, they all follow the same principle of taking a reference image to guide your prompt. This is because sometimes a prompt is not enough and it takes a reference image to fully give Arcana the clarity of your intended creation.
Arcana AI offers a host of powerful models each trained on their own unique datasets and styles. A common misperception is to rule out a model after prompting without at least conjuring 4 images using the same prompt and model.
If after 4 image results are still not to your liking, then one should either switch models better suited to your creative needs or adjust your prompt.
But how do you know which model to even begin with when there are so many? Well, dear readers, there is an "i" on the top right corner of each model allowing you to see the strengths and weaknesses of each model as well as view images made from the model so you can assess its creative style:
You can even take prompts from inside the info modal by clicking reconjure and editing the prompt to your liking:
For a brief review of the models let's use a basic and vague prompt "Wizard, turtle"
PLEASE NOTE: we have not selected any styles in this demonstration and selecting a style can radically improve or change the quality of the prompt depending on the model you use. Also, some models may appear to be photorealistic but if you simply add to the prompt "PIXAR style" or "Anime image of..." the model with reveal itself to be much broader in style than initially meets the eye with this test. For greater clarity, us the "info" button on the top right of each model.
If you fall in love with an image or a prompt that is really working for you, don't worry the prompt is always available in the metadata of the image:
Simply click save prompt to add it to your collection of saved prompts.
To access or edit your collection of saved prompts navigate to the right drawer above the models list:
Arcana boasts over 70+ styles!
You can select up to 2 styles. Mixing and matching styles creates very interesting outputs. If you see a style with the word XL after it. Such as CYBORG XL it is because it was made specifically for one of the XL models. (You will know if a model is an XL model b/c its name will always be followed with XL. For examples IMAGINE XL or NIJI XL. ) If you select and XL style and use a non XL model it will still work but you will not gain the same benefits had you selected and XL model. See the style guide for a breakdown of every style. (coming soon)
To learn about each tool see
