Stable Diffusion 3 Medium with the ability to run efficiently on user laptops published by Stability AI

Stability AI on Wednesday released a smaller version of its Stable Diffusion 3 (SD3) artificial intelligence (AI) model. Called Stable Diffusion 3 Medium, the company introduced a smaller text-to-image model as its most advanced image generation model. While retaining all the functionality of the larger generative AI model, the latest tool has lower GPU requirements and consumes less power than previous models. Open weights are also available on Hugging Face, and the company says this AI model can work effectively on PCs and laptops.

Stability AI introduces Stable Diffusion 3 Medium

While the Stable Diffusion 3 model (now called the Stable Diffusion 3 Large) became publicly available in April, its high GPU and computing requirements made it difficult for most people to work efficiently with a consumer-grade PC or laptop. The company solves this problem by offering Stable Diffusion 3 Medium, which can run on most laptops and PCs.

According to a report by VentureBeat, the minimum requirement for the AI ​​model is 5GB of GPU VRAM, and the recommended requirement is 16GB of GPU VRAM. Namely, the Nvidia GeForce RTX 3090 has 24 GB of GDDR6X VRAM.

Despite its smaller size of two billion parameters (as opposed to SD3 Large’s eight billion parameters), Stability AI said in a press release that Stable Diffusion 3 Medium will be able to demonstrate a similar level of efficiency to its larger counterpart. The latest image generation model will provide detailed photorealistic prints as well as high quality prints in flexible styles. To improve realism in hands and faces, the AI ​​company uses a 16-channel VAE (variational autoencoder).

Fast adherence will also be on par with the larger model. SD3 Medium can understand complex instructions involving spatial thinking, compositional elements, actions and styles. Furthermore, typography, which was a common pitfall of image generation models, has also been improved in the latest AI model, the company added.

Stable Diffusion 3 Medium has become generally available through the AI-powered Fireworks API (App Programming Interface). The text-to-image AI model can also be accessed via the Stable Assistant platform or the Stable Artisan Discord server. Other open weights are also available under a non-commercial license from Hugging Face. In order to use it for commercial purposes, users will need to obtain a creator license from the company.