©MeSomewhere aims to provide a simplified, reliable, and free way to use state-of-the-art image generation models for common image creation and editing tasks. We accomplish this by using a single set of pretrained latent diffusion parameters for all tasks, rather than using a separate set of fined-tuned parameters for each task. ©MeSomewhere takes inspiration from the latest research in latent diffusion models, as well as implements custom unpublished processing techniques (we intend to publish and open-source these techniques in the near future). Please see the list of projects and publications below for notable sources of inspiration

This application is still in active development. Please email any comments or suggestions to img23d@meztech.io. Happy generating!