ApX logoApX logo

FLUX.2 Klein 9B

Parameters

9B

Modality

Image

Auxiliary Parameters

80M

Architecture

Diffusion Transformer

License

Apache 2.0

Release Date

10 Feb 2025

API Pricing (per 1M)

Self-hosted only

Hardware Sizing

Precision

Workload Memory: 7.19 GB

INT4 · 1024x1024

DiT Backbone Weights: 4.19 GB

VAE Decode Workspace: 0.50 GB

Runtime & Graphs: 2.50 GB

Minimum Required Hardware

1x NVIDIA L4 24GB

7.2 / 21.6 GB usable

1x RTX 3060 12GB

7.2 / 10.8 GB usable

Apple M4 24GB

7.2 / 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: 12.3k · Text: 80M enc · Timestep (t)x 32 DiT blocksAdaLN ModulationNorm · Timestep scale/shiftJoint / Cross-Attention32 headsHead dim: 128+AdaLN ModulationNorm · Timestep scale/shiftFeed-Forward (MLP)Activation+Final Norm (AdaLN Scale)Predicted Noise / LatentsDenoised Latents to VAE

About FLUX.2 Klein 9B

FLUX.2 Klein 9B is an efficient 9B parameter DiT balanced for intermediate workstation and workstation GPUs, offering high visual fidelity while reducing activation and weight footprints.

Technical Specifications

Attention

Attention Heads

32

Attention Head Dimension

128

Dimensions

Auxiliary Parameters

80M

Hidden Dimension Size

12,288

Number of Layers

32

About Flux 2


Other Flux 2 Models