Stable Diffusion 3.5 Large
Open Weights
Stability AI
Stability AI's 8-billion-parameter open text-to-image model (October 2024), the largest of the SD 3.5 family. Built on the Multimodal Diffusion Transformer (MMDiT) architecture, it improves prompt adherence and image quality over SDXL and is released under the permissive Stability Community License.
Strengths
- 8B open text-to-image model on the MMDiT architecture
- Improved prompt adherence and quality over SDXL
- Permissive Stability Community License (free under $1M revenue)
- Large open ecosystem (ComfyUI, fine-tunes, ControlNets)
- Runs on consumer/prosumer GPUs
Caveats
- Surpassed by FLUX at the open-weight quality frontier
- Community License requires enterprise terms above $1M revenue
- Text rendering trails specialist models
- Vendor
- Stability AI
- License
- Stability Community License
- Release Date
- 2024-10-22
- Modalities
- textimage
Capabilities
Vision
Audio
Video
Tool Use
Resources
Pricing
Open weights under the Stability Community License - free for research and for organizations under $1M revenue; enterprise license above that. Also via the Stability API.