From the model card
Fields below are copied from the tags and counters on the HuggingFace repository stabilityai/stable-diffusion-3.5-medium at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- stabilityai
- Pipeline tag
- text-to-image
- Library
- Diffusers
- Weight formats
- safetensors
- License tag
other— read the license file in the repo before relying on it- Language tags
- English (en)
- Papers cited
- arXiv:2403.03206
- Downloads (HF counter at last fetch)
- 616,212
- Likes (HF counter at last fetch)
- 1,007
- Model card
- https://huggingface.co/stabilityai/stable-diffusion-3.5-medium
Use cases
- High-quality text-to-image generation with improved text rendering
- Artistic image generation where prompt adherence matters
- Faster inference alternative to SD3.5-Large for prototyping
- Fine-tuning base for style-specific image generation
Pros
- MMDiT architecture notably improves text legibility in generated images
- Faster than SD3.5-Large while maintaining reasonable quality
- Diffusers pipeline compatible — standard tooling works
- Better prompt adherence than SDXL on complex compositions
Cons
- License restricts commercial use for organizations above revenue threshold — check terms
- Medium variant underperforms Large on photorealism and fine detail
- Requires more VRAM than SDXL models of similar output resolution
- Slower than SDXL Turbo/Lightning distillations at generation time
Tags
diffuserssafetensorstext-to-imagestable-diffusionenarxiv:2403.03206license:otherregion:us