ApX logoApX logo

Stable Diffusion 3 Medium

Parameters

2B

Modality

Image

Auxiliary Parameters

6.16B

Architecture

Diffusion Transformer

License

Stability AI Community License

Release Date

12 Jun 2024

API Pricing (per 1M)

Self-hosted only

Hardware Sizing

Precision

Workload Memory: 3.57 GB

INT4 · 1024x1024

DiT Backbone Weights: 0.93 GB

VAE Decode Workspace: 0.14 GB

Runtime & Graphs: 2.50 GB

Minimum Required Hardware

1x NVIDIA L4 24GB

3.6 / 21.6 GB usable

1x RTX 3060 12GB

3.6 / 10.8 GB usable

Apple M4 24GB

3.6 / 18.0 GB usable

Determine requirements for your specific workload:

Open in VRAM Calculator

Architecture Diagram

Noisy Image LatentsLatent Image Space (VAE)Patch Embedding2D Latent PatchifyHidden: 4.1k · Text: 6.16B enc · Timestep (t)x 24 DiT blocksAdaLN ModulationNorm · Timestep scale/shiftJoint / Cross-Attention24 headsHead dim: 64+AdaLN ModulationNorm · Timestep scale/shiftFeed-Forward (MLP)Activation+Final Norm (AdaLN Scale)Predicted Noise / LatentsDenoised Latents to VAE

About Stable Diffusion 3 Medium

Stable Diffusion 3 Medium is the pioneering 2.0B parameter MMDiT release from Stability AI, decoupling visual and textual token representations across separate transformer streams.

Technical Specifications

Attention

Attention Heads

24

Attention Head Dimension

64

Dimensions

Auxiliary Parameters

6.16B

Hidden Dimension Size

4,096

Number of Layers

24

About Stable Diffusion 3


Other Stable Diffusion 3 Models
  • No related models available