CompVis logo

Stable Diffusion

Stable Diffusionvv1Current
byCompVisCompVis(open source)
Released August 22, 2022
Context--
CategoryImage Model

About Stable Diffusion

Stable Diffusion is a latent text-to-image diffusion model from CompVis, developed with compute support from Stability AI and support from LAION. The v1 model generates and modifies images from text prompts using a latent diffusion architecture and a frozen CLIP ViT-L/14 text encoder.

Capabilities

text to imageimage generationimage to imageresearch

Input Modalities

textimage

Output Modalities

image

Technical Details

API Identifier
CompVis/stable-diffusion-v1
Category
Image Model

Tags

text-to-imagediffusionlatent-diffusionimage-generationopen-weightsresearchcreative-ai

Benchmarks

Performance scores for Stable Diffusion across standard benchmarks.

CVPR 2022 latent diffusion paperhttps://openaccess.thecvf.com/content/CVPR2022/html/Rombach_High-Resolution_Image_Synthesis_With_Latent_Diffusion_Models_CVPR_2022_paper.html · Jun 2022
%

Competing Models

Same pricing tier — direct alternatives to Stable Diffusion

specialty
CompVis logo
CompVisopen source

Computer vision and generative AI research from LMU Munich

Founded 2017Munich, Germany
View full profile