FLUX - Image Generation Models

Amazing Image Generation Models from the Black Forest.

Visit Website
FLUX - Image Generation Models

Introduction

Introduction


Open Source Text-to-Image Models

Black Forest Labs has developed Flux, the largest open source text-to-image model to date, with 12B parameters. It is available on fal along with three variations: FLUX.1 [dev], FLUX.1 [schnell], and FLUX.1 [pro].

Image Generation Model

FLUX.1 is a new open-source image generation model developed by Black Forest Labs.

Features


Enhanced Image Quality

Flux models generate stunning visuals at higher resolutions, achieving highly realistic and anatomically accurate images with improved prompt adherence.

Advanced Human Anatomy and Photorealism

Flux models deliver aesthetics reminiscent of Midjourney, providing highly realistic and anatomically accurate images.

Improved Prompt Adherence

Flux models generate more accurate and relevant images based on user inputs.

State-of-the-art Performance

FLUX.1 models have state-of-the-art performance in prompt following, visual quality, image detail, and output diversity.

Text Handling

FLUX.1 can handle tricky words with repeated letters, making it great for designs where text needs to be accurate.

Complex Composition

FLUX.1 is amazing at following complex instructions about where things should go in an image.

Hand Rendering

FLUX.1 does a pretty good job at rendering hands, with the right number of fingers in the right places.

Exceptional Speed

Flux Schnell, a distilled version of the base model, operates up to 10 times faster than the base model, making it ideal for high-demand applications.

Usage


Inference Engine Integration

fal's cutting-edge inference engine allows Flux models to run up to 2x faster than with eager torch, resulting in faster processing times while maintaining exceptional quality and detail.

AI Model Upscaler

Black Forest Labs has released AuraSR v2, an open-source single step GAN upscaler based on the Adobe Gigagan paper, which excels in upscaling images generated by text-to-image models.

Price


Model Variants

FLUX.1 comes in three variants: FLUX.1 [pro], FLUX.1 [dev], and FLUX.1 [schnell], each with different performance and pricing options.

Pricing

FLUX.1 is priced per image, with different rates for each variant.

Traffic Data

Traffic Overview

Monthly Visits
330,345
Bounce Rate
50.34%
Pages Per Visit
2.02
Visit Duration
00:01:06
Global Rank
163,473
Country Rank
150,022 (US)

Visits Over Time

Traffic Sources