Stable Diffusion
Stable Diffusion is a text-to-image diffusion model trained on 512x512 images from the LAION-5B dataset that can generate images from text prompts using latent diffusion techniques. The model has been widely adopted and optimized across multiple platforms, including consumer GPUs with as little as 3.2GB VRAM and Apple M1/M2 chips, while supporting various extensions such as image-to-image generation, inpainting, and textual inversion through the Diffusers library. Recent developments include fine-tuning capabilities via reinforcement learning methods like DDPO, language-specific variants such as Japanese Stable Diffusion, and acceleration through tools like ONNX Runtime.
Updated 3 August 2026
Specifications
No specifications recorded yet.
Latest developments
Multimodality and Large Multimodal Models (LMMs)
Chip Huyen · 2 years ago ·
47
Accelerating over 130,000 Hugging Face models with ONNX Runtime
Hugging Face Blog · 2 years ago ·
15
Finetune Stable Diffusion Models with DDPO via TRL
Hugging Face Blog · 2 years ago ·
50
Text-to-Image: Diffusion, Text Conditioning, Guidance, Latent Space
Eugene Yan · 3 years ago ·
14
Japanese Stable Diffusion
Hugging Face Blog · 3 years ago ·
26
What's new in Diffusers? 🎨
Hugging Face Blog · 3 years ago ·
45
Stable Diffusion with 🧨 Diffusers
Hugging Face Blog · 3 years ago ·
41
Q4 2023
- Multimodality and Large Multimodal Models (LMMs)
- Accelerating over 130,000 Hugging Face models with ONNX Runtime
Q3 2023
Q4 2022
Q3 2022
Relationships
Products & technology
- Integrated with Diffusers · 1 source
- Derived from LAION-5B · 1 source
- Stability AI develops this model · 1 source
- Japanese Stable Diffusion derived from this model · 1 source
- TRL integrated with this model · 1 source