ApX logoApX logo

FLUX.1 Dev

Parameters

11.9B

Modality

Image

Auxiliary Parameters

4.9B

Architecture

Diffusion Transformer

License

FLUX.1-dev Non-Commercial License

Release Date

1 Aug 2024

API Pricing (per 1M)

Self-hosted only

Hardware & System Requirements

Precision

Native

Workload Memory: 8.71 GB

INT4 · 1024x1024

DiT Backbone Weights: 5.54 GB

VAE Decode Workspace: 0.67 GB

Runtime & Graphs: 2.50 GB

Minimum Required Hardware

1x NVIDIA L4 24GB

8.7 / 21.6 GB usable

1x RTX 3060 12GB

8.7 / 10.8 GB usable

Apple M4 24GB

8.7 / 18.0 GB usable

Determine requirements for your specific workload:

Open in VRAM Calculator

Runtime requirements

Checkpoint format

FP16

CUDA

>= 11.8

PyTorch

>= 2.1

Transformers

>= 4.40.1

ArchitecturePrecisionAttention kernel

Blackwell, Hopper

Native FP16

FlashAttention-3

Ada Lovelace, Ampere

Native FP16

FlashAttention-2

Architecture Diagram

About FLUX.1 Dev

FLUX.1 Dev is a 12B parameter guidance-distilled rectified flow model by Black Forest Labs, delivering competitive photorealism, typography rendering, and complex composition adherence with multi-modal dual attention.

Specifications

Attention

Attention Heads

24

Attention Head Dimension

128

Dimensions

Base Precision

FP16

Auxiliary Parameters

4.9B

Hidden Dimension Size

4,096

Number of Layers

57