What is Stability AI (Stable Diffusion)?

Stability AI is an advanced AI model that synthesizes billions of images from text prompts. It incorporates research from latent diffusion models similar to those used by Dall-E and Google Imagen. This model was initially introduced by Katherine Crowson from the CompVis team.

The art generated by Stability AI is trained on Aesthetics, a specific subset of the LAION 5B dataset. This technology refines images from LAION-5B and translates the ratings into the CLIP model. It is capable of rendering images at 512x512 pixels in mere seconds using standard consumer GPUs.

For further insights on Stability AI, consider exploring the following resources: - Stability AI GitHub - Stability AI Download (Website)

How Does the Stability AI Generator Operate?

The Stability AI generator allows users to create images from a library of over 5 billion visuals, utilizing 800 million parameters. It operates effectively with 10 Gb VRAM, and you can test it for free in beta on Discord. However, each prompt must be unique to avoid generating identical images.

To access Stability AI for free, you need to join a 'Waitlist' at the designated address. You can expect an email notification within 1-2 weeks if your application is approved, allowing you to log in via a Discord invitation link included in the email.

Once you join the Discord server, there may be a brief wait before you can create images. It’s essential to have a verified Discord account. Once approved, you can issue one command every 60 seconds in the ‘#dream’ chat room after the waiting period.

Here’s a summary of the commands available for use in the ‘#dream’ chat room:

usage: !dream [-h] [--tokenize] [--height HEIGHT] [--width WIDTH]

[--cfg_scale CFG_SCALE] [--number NUMBER]

[--separate-images] [--grid] [--sampler SAMPLER]

[--steps STEPS] [--seed SEED] [--prior PRIOR]

[--ascii] [--asciicols ASCIICOLS] [prompt ...]

positional arguments:

prompt

optional arguments:

-h, --help

—tokenize, -t show CLIP tokenization output

--height HEIGHT, -H HEIGHT
	[512] height of image (multiple of 64)

—width WIDTH, -W WIDTH [512] width of image (multiple of 64)

--cfg_scale CFG_SCALE, -C CFG_SCALE
	[7.0] CFG scale factor

—number NUMBER, -n NUMBER [1] number of images

--separate-images, -i
	Return multiple images as separate files.

—grid, -g Composite multiple images into a grid.

--sampler SAMPLER, -A SAMPLER
	[k_lms] (ddim, plms, k_euler, etc.)

—steps STEPS, -s STEPS [50] number of steps

--seed SEED, -S SEED
	random seed to use

—prior PRIOR, -p PRIOR vector_adjust_prior

--ascii, -a

a trip back to 1970

—asciicols ASCIICOLS, -ac ASCIICOLS If ascii, number of text columns

If you enjoy content related to artificial intelligence, check out some of my most popular articles: - 30 Different Images Created by DALL-E 2, DALL-E Mini, and Midjourney - How Satisfied Are You With The Latest Update for Dall-E 2 - OpenAI’s DALL-E 2 AI: Concerns for Artists

Top 20 Images Generated by Stability AI

Before diving into the images produced by Stability AI, it's important to note that each image was created using the ‘!dream’ command, with the specific prompts provided below.

Additionally, I included three images from Dall-E 2 for comparison with other AI technologies, giving you a clearer perspective. Let's take a look at the ‘!dream’ command used as an example:

!dream "the ufo, high saturation, high contrast, vibrant " -H 640 -C 20.0 -n 4 -i -s 150 -S 788846748

Prompt: Anthropomorphic goose wearing steampunk leather goggles, leather facial mask, a long leather coat, dramatic lighting, full body shot, Sigma 50mm

Prompt: Iron Man and Howard the Duck, various digital art

Prompt: 35mm photo of Elizabeth from Bioshock Infinite, cosplay, digital art.

Prompt: Elon Musk (1987), horror, game design fanart.

Prompt: Film still of a Squirrel working in a research lab filling test tubes, 8k.

Prompt: Highly detailed portrait of Okabe, Rintarou, in Steins;Gate, fantasy art by various artists.

Prompt: Wearing Warhammer 40k champion black armor, anthropomorphic Shiba Inu, stunning 3D render.

Prompt: Little demon building a sandcastle, family picture.

Prompt: A beautiful painting of a beautiful painting of a beautiful painting of a beautiful painting.

Prompt: Saria from Zelda in anime style.

Prompt: Saria as an emo girl.

Prompt: A highly detailed futuristic mechanical lady, seductive confident pose.

Prompt: Venomized Elmo, nightmarish, intricate.

Prompt: Couple running beside a river, steep riverbank, white clothes, happy summer days.

Prompt: Stylized portrait of an artistic pose, young cute serious lady.

Prompt: Perfectly centered close-up portrait, zombie pirate, highly detailed.

Prompt: Portrait of Dr. John A. Zoidberg, photorealistic studio lighting.

Prompt: Portrait of a Bill Gates as a Jedi commander, wearing tactical gear.

Prompt: An epic flowering alien landscape in the style of origami.

Prompt: Photo of a person who lost the thrill of living long ago, but life goes on.

If you enjoyed this content, don't forget to give it a clap! Your feedback helps me create high-quality material for you. If you're interested in more articles in this series, feel free to leave a comment!