ModelCards

Stable Diffusion 3.5 Large

Image modelOpen weightsCard summarized

At 8 billion parameters, with superior quality and prompt adherence, this base model is the most powerful in the Stable Diffusion family.

Source: Stable Diffusion 3.5 Large model card (Hugging Face), SD 3.5 launch post and API reference

Use it via API

$0.065 / image
Code exampleCompare

Where to call it

  • Stability AI Platform APIsd3.5-large
  • Self-host (weights)stabilityai/stable-diffusion-3.5-large
  • Amazon Bedrock

Supports

  • Image input
Full feature support

Good for

  • Generating artwork and use in design and other artistic processes
  • Educational or creative tools
  • Research on generative models and their limitations

Watch out for

Overview

Stability AI · 03Large, front

SD 3.5 Large

At 8 billion parameters, with superior quality and prompt adherence, this base model is the most powerful in the Stable Diffusion family.

Typical price
$0.065 / image
SD 3.5 Large, 1 MP image
Max resolution
1 MP via API
Aspect ratios
9 options
Modes
text-to-image, image-to-image
Released
22 Oct 2024

Stable Diffusion 3.5 Large is Stability AI's largest open-weights image model: an 8.1-billion-parameter Multimodal Diffusion Transformer (MMDiT) released on 22 Oct 2024 with a distilled 4-step Large Turbo. Stability says it leads in prompt adherence and suits professional use at 1 megapixel. It also powers Stable Image Ultra. Weights are on Hugging Face under the Stability AI Community License, and both models are served on Stability's API. model card; launch post; release notes

sd3.5-large
Typical price
$0.065 / image
Max resolution
1 MP via API (default 1024×1024)
Aspect ratios
21:9, 16:9, 3:2, 5:4, 1:1, 4:5, 2:3, 9:16, 9:21
Weights
Open · Stability AI Community License (free under $1M annual revenue; Enterprise License above)
Released
22 Oct 2024
  • text-to-image
  • image-to-image