DreamStudio beta Updates 1-Dec 22
The aim of DreamStudio beta is to share the incredible technology of diffusion models with the entire world by making it accessible and affordable with the absolute fastest compute available.
Our ongoing research and optimization efforts allow us to pass our efficiency on to our customers -- today we're announcing an 80% reduction in price for standard image generation (from 1.0 credits to 0.2 credits.)
This change is in immediate effect for all DreamStudio and API users.
Calculation data and more details are available on our Platform site - platform.stability.ai
We are pleased to announce that Stable Diffusion v2.0 and v2.0 - 768 are available now on DreamStudio beta as an optional model in the dropdown menu tool at the bottom right of the parameters panel, along with the editing model option Stable Diffusion Inpainting v2.0.
The base Stable Diffusion v2.0 is trained on a 512x512 dataset. The larger Stable Diffusion v2.0 - 768 model is trained on a higher resolution dataset of 768x768 and supports greater coherence at higher native resolutions such as 1024x576 and 1024x768.
Future updates are planned for DreamStudio beta and progress is underway! We’re excited to bring you more feature updates in the weeks ahead, along with events on our Discord, Twitch, YouTube and Instagram channels, along with some other surprises!
Visit beta.dreamstudio.ai to create your free account and start dreaming with text-to-image, img2img, inpainting and detailed control.
Join our 100k member community on Discord and share tips, tricks and prompting secrets with fellow latent space cadets.