models/Stable Diffusion 3.5 Large - Large

Stable Diffusion 3.5 Large - Large

7/2/2025

5:19:11 AM

Related Keywords & Tags

base model,checkpoint,large,multimodal diffusion transformer,qk-normalization,sd 3.5 large,sd3.5,stability ai,stable diffusion,stable diffusion 3.5,stable diffusion 3.5 large,text-to-image generative model,theally

An abstract face emerging from a swirling vortex of midnight black, depicted with chaotic scribbles and minimalist lines illuminated by soft, shimmering white light.

A digital pointillist illustration depicting a little girl in a flowing white dress sitting beside a black cat on a hill, set against a monochrome sunset with a textured star-like dot pattern.

Legs clad in black tights wearing vibrant red stiletto heels with ankle straps, set against a dynamic abstract background with gold leaf and brushstroke textures.

A solitary woman stands in a field illuminated by golden hour light during a dramatic sunset, surrounded by swaying grasses and distant jagged mountains in an ethereal, abstract expressionist style.

A tall skeletal tower made of corroded metal stands alone on a cracked, flat ground. The scene has a minimalist, impressionistic style with soft white light highlighting the tower's linear shapes against a dark background.

Monochrome silhouette of a Parisian flâneur with a hat and cane on a cobblestone street, framed by minimalist Haussmann architecture and a tree.

Minimalist digital illustration of a solitary blonde woman with sleek straight hair seated on a chair against a deep blue background, holding a vibrant fuchsia coffee cup.

A couple walking closely under a dark umbrella during a rainstorm at sunset, with rain droplets on a wet window and a blurred textured background.

Abstract artistic depiction of gingko leaves swirling in vivid autumn colors like crimson, gold, and ochre on textured washi paper.

A small, delicate robot with smooth edges stands in a misty scene filled with soft pink cherry blossoms, evoking the style of a traditional Japanese woodblock print.

A solitary female warrior with pale lavender skin stands wearing futuristic armor, painted with bold watercolor brushstrokes in black, white, and red on a minimalist lavender background.

A solitary young figure in silhouette stands near a symbolic plant against a minimalist sky with a warm sunset gradient, emphasizing peaceful defiance and introspective freedom.

Recommended Prompts

a detailed illustration of a cute rainbow colorful boba tea sitting in a window sill, tiki bar looking out onto a tropical oceanfront background, shallow depth of field, cinematic lighting, dappled sunlight, lens flare bokeh

Recommended Parameters

samplers

DPM++ 2M

steps

cfg

4.5

Creator Sponsors

Please see our Quickstart Guide to Stable Diffusion 3.5 for all the latest info!

Stable Diffusion 3.5 Large is a Multimodal Diffusion Transformer (MMDiT) text-to-image model that features improved performance in image quality, typography, complex prompt understanding, and resource-efficiency.

Please note: This model is released under the Stability Community License. Visit Stability AI to learn or contact us for commercial licensing details.

Model Description

Developed by: Stability AI
Model type: MMDiT text-to-image generative model
Model Description: This model generates images based on text prompts. It is a Multimodal Diffusion Transformer that use three fixed, pretrained text encoders, and with QK-normalization to improve training stability.

License

Community License: Free for research, non-commercial, and commercial use for organizations or individuals with less than $1M in total annual revenue. More details can be found in the Community License Agreement. Read more at https://stability.ai/license.
For individuals and organizations with annual revenue above $1M: please contact us to get an Enterprise License.

Implementation Details

QK Normalization: Implements the QK normalization technique to improve training Stability.
Text Encoders：
- CLIPs: OpenCLIP-ViT/G, CLIP-ViT/L, context length 77 tokens
- T5: T5-xxl, context length 77/256 tokens at different stages of training
Training Data and Strategy:
This model was trained on a wide variety of data, including synthetic data and filtered publicly available data.