On Thursday, Stability AI announced Stable Diffusion 3, an open-weights next-generation image-synthesis model. It reportedly generates detailed, multi-subject images with improved quality and accuracy in text generation.
Stability says that its Stable Diffusion 3 family of models range in size from 800 million to 8 billion parameters, accommodating different devices and model capabilities. Larger models need more VRAM for operation.
[
add
]
[
|
|
...
]