What makes FLUX different from Stable Diffusion or Midjourney?

FLUX uses a hybrid multimodal and parallel diffusion transformer architecture with flow matching rather than traditional DDPM diffusion, enabling faster generation. It offers a 12B parameter open-weight model (FLUX.1 [dev]) — larger than most open alternatives — with superior prompt adherence, better text rendering, and both open-source and commercial API tiers.

How do I access FLUX.1 models?

You can access open-weight models (FLUX.1 [schnell] and [dev]) on Hugging Face and GitHub for free. Proprietary Pro models are available via the BFL API at bfl.ai or through partner platforms including fal.ai, Replicate, Together AI, Adobe Firefly, and Freepik.

Is there a free way to try FLUX?

Yes. The BFL Playground at bfl.ai lets users try models without writing code. FLUX.1 [schnell] and FLUX.2 [klein] are available as open-weight downloads for free local use. Third-party platforms like Freepik and Replicate also offer limited free trials.

FLUX.1 AI

Freemium🇩🇪🇩🇪 GermanyUnicornHigh Burn

Frontier image generation from the creators of Stable Diffusion

Image Gen Image Gen Gen AI Multimodal AI Graphic Design

Overall score

Heat score

Claim Tool Suggest Edits

Pricing

FLUX.1 [schnell] (Open)Free (Apache 2.0)

FLUX.1 [dev] (Open)Free (non-commercial)

FLUX1.1 [pro] API$0.04/image

FLUX1.1 [pro] Ultra / Raw$0.06/image

FLUX.1 Kontext [pro]$0.04/image

FLUX.1 Kontext [max]$0.08/image

FLUX.2 [pro]from $0.03/image

FLUX.2 [klein] 4Bfrom $0.014/image

Purchase Subscription

Technical Specs

Inputs

Text Prompt, Reference Image, Style Image, Depth Map, Edge Map, Inpainting Mask

Outputs

AI Generated Image, Edited Image, Inpainted Image, High-Resolution Output, Photorealistic Render

AI Type

Image Generation

Model Architecture

Diffusion

Daily Prompts

N/A

Context Length

N/A

Output Quality

Accuracy

93%

Content

91%

Reasoning

78%

Company Profile

Company

Black Forest Labs

Founded

2024

Freiburg, Baden-Württemberg, Germany

Employees

150

Total Raised / Total Funding

$450M

Revenue

N/A

Valuation

$3.25B

ARR

N/A

CEO

Robin Rombach

Overview

Estimated Paid Users

N/A

Current estimate

Total Earnings Till Date

N/A

+12.31% from last month

Market Share

Current share

Average Session

Per active user

Hallucination Rate

18%

Model quality signal

Growth Rate

+8.00%

Monthly active users

Burn Rate

N/A

Total expenses / years active

Paid User Gain

+36.36%

Monthly paid user trend

No demo video available yet.

Platforms

Web App

Available

References

VentureBeat: Black Forest Labs launches FLUX.1, secures $31M seed

•

venturebeat.com

TechCrunch: Black Forest Labs releases an API (Oct 2024)

•

techcrunch.com

TechCrunch: Black Forest Labs raises $300M at $3.25B valuation

•

techcrunch.com

GlobeNewswire: Series B announcement press release

•

globenewswire.com

Salesforce Ventures: Investment thesis for BFL

•

salesforceventures.com

Northzone: Investment in Black Forest Labs

•

northzone.com

Wikipedia: FLUX (text-to-image model)

•

en.wikipedia.org

Sacra: Black Forest Labs revenue, valuation and funding

•

sacra.com

Artificial Analysis: FLUX model family analysis

•

artificialanalysis.ai

TNGlobal: Temasek backs Black Forest Labs Series B

•

technode.global

BFL Official Blog: $300M Series B announcement

•

bfl.ai

DeepLearning.AI The Batch: FLUX.1 outperforms top text-to-image models

•

deeplearning.ai

Profit Analysis

-$138.5M

Total Loss

$172.2M

Total Profit

Performance Metrics

Accuracy

93%

Context

91%

Reasoning

78%

Safety

72%

Benchmarks

Custom Internal Benchmark

100%

FLUX.1 AI Models

FLUX.1 [schnell]

Type: Image

Description: Fastest open-weight model, 12B parameters, Apache 2.0 license for commercial and personal use. Optimized for speed over maximum quality.

Architecture: Diffusion

FLUX.1 [dev]

Type: Image

Description: 12B parameter open-weight distillation of FLUX.1 [pro]. Non-commercial license. Top-ranked on Hugging Face with 500K+ downloads at launch.

Architecture: Diffusion

FLUX.1 [pro]

Type: Image

Description: Proprietary closed-source model available via API. Highest quality at launch, $0.05/image initial pricing. Superseded by FLUX1.1 [pro].

Architecture: Diffusion

FLUX1.1 [pro]

Type: Image

Description: Flagship proprietary model at $0.04/image. 6× faster than FLUX.1 [pro], improved prompt adherence and image diversity, up to 1440px. ELO 1153 at launch.

Architecture: Diffusion

FLUX1.1 [pro] Ultra

Type: Image

Description: Ultra-high-resolution variant supporting up to 4MP images at $0.06/image. Maintains ~10s generation time despite 4× resolution increase.

Architecture: Diffusion

Funding Rounds & Investors

Total Funding

$450M

Rounds

Series B

$300M

Dec 2025

Co-led by AMP and Salesforce Ventures at $3.25B post-money valuation; total funds raised exceed $450M; source: GlobeNewswire press release 2025-12-01 and TechCrunch 2025-12-01

Series A

N/A

Jun 2025

Previously unannounced Series A led by Andreessen Horowitz; disclosed retroactively at time of Series B announcement December 2025; exact amount undisclosed; source: GlobeNewswire 2025-12-01

Seed

$31M

Aug 2024

Led by a16z; emerged from stealth simultaneously with FLUX.1 model launch; source: VentureBeat 2024-08-01 and TechCrunch 2024-10-03

Founders/Team

Robin Rombach

Co-Founder & CEO

Andreas Blattmann

Co-Founder & CTO

Patrick Esser

Co-Founder

Dominik Lorenz

Co-Founder

Direct competitors

No direct competitors available.

Change Log / Major Updates

2025 · Jun 1

FLUX.1 Kontext introduced a new model series enabling in-context image generation and editing from combined text and image prompts. Available in Max ($0.08), Pro ($0.04), and open-weight Dev variants. Adobe integrated FLUX.1 Kontext [pro] into Photoshop Generative Fill.

2025 · Jul 31

BFL released FLUX.1 Krea Dev, a model developed in partnership with Krea AI focused on better aesthetic diversity, more varied styles, and improved realism compared to base FLUX.1 [dev].

2025 · Nov 25

FLUX.2 introduced four models with multi-reference support (up to 10 images), HEX color matching for brand precision, improved text rendering, and the open-weight Apache 2.0 Klein series (4B and 9B) offering sub-second generation from $0.014/image.

Compliance, Integrations & Support

Industry: Not specified

Compliances: Not specified

Integrations: Adobe Photoshop, Canva, Meta, Mistral AI Le Chat, fal.ai, Replicate, Together AI, Freepik, Microsoft Azure AI Foundry, Hugging Face, GitHub, Picsart, ElevenLabs, VSCO, Vercel, Krea AI, xAI Grok, Snap, Shutterstock, NVIDIA Blackwell

Support:email, help center, community forum, enterprise support, api documentation

Target audience: Developers, Creative Professionals, Designers, Enterprise Teams, Researchers, Content Creators, Marketing Agencies, App Builders

Supported languages: English, German, French, Spanish, Japanese, Chinese, Italian, Portuguese

FLUX.1 AI Acquisitions

No acquisition records available.

Reviews & Rating

0 reviews

No reviews yet

Be the first to share how FLUX.1 AI performs for your workflow.

0.0

Accuracy

0.0

Ease of Use

0.0

Output Quality

0.0

Security

0.0

Social Feed

No social feed available for this tool yet.

More About FLUX.1 AI

In the spring of 2024, three researchers walked out of Stability AI carrying something more valuable than any equity stake: the intellectual architecture of modern image generation itself. Robin Rombach, Andreas Blattmann, and Patrick Esser — the scientists behind latent diffusion and Stable Diffusion — had watched their previous employer struggle to commercialize one of AI's most consequential breakthroughs. They decided to try it themselves, properly this time.

What Is FLUX.1?

FLUX.1 is a family of text-to-image generation models built on a novel hybrid multimodal and parallel diffusion transformer architecture with 12 billion parameters. Unlike earlier diffusion systems, FLUX uses flow matching — a more direct path from noise to image — enabling generation in as little as 4.5 seconds at industry-leading quality. Within days of its August 2024 launch, FLUX.1 [pro] and FLUX.1 [dev] claimed the top two spots on the Artificial Analysis Text-to-Image Arena Leaderboard, a human-preference benchmark with tens of thousands of blind comparisons. FLUX1.1 [pro] later hit an ELO of 1153, surpassing Midjourney 6.1 (1100) and Ideogram v2 (1108).

The Model Family

FLUX.1 [schnell] — Apache 2.0 open-weight model, fastest generation, commercial use permitted
FLUX.1 [dev] — 12B parameter open-weight distillation, non-commercial license, top-ranked on Hugging Face
FLUX.1 [pro] / FLUX1.1 [pro] — proprietary API models; FLUX1.1 [pro] is 6× faster than its predecessor at $0.04/image
FLUX.1 Tools — Fill (inpainting/outpainting), Depth, Canny, Redux editing suite released November 2024
FLUX.1 Kontext — in-context image editing with Max, Pro, and Dev tiers
FLUX.2 — next-gen series with multi-reference support, HEX color matching, and sub-second generation via the open-weight Klein models

Traction and Enterprise Adoption

FLUX models have accumulated 400 million+ downloads across Hugging Face, fal.ai, Replicate, and Together AI. Enterprise partners include Adobe (Photoshop Generative Fill uses FLUX.1 Kontext [pro]), Canva, Meta, Mistral AI, Picsart, ElevenLabs, VSCO, Vercel, and Freepik. A reported $140M agreement with Meta and approximately $300M in enterprise contract value underscores the commercial weight behind the research credentials. By December 2025, the company had closed a $300M Series B at a $3.25B valuation, bringing total funding to over $450M.

FLUX.1 AI FAQ's

What is FLUX.1 and who makes it?

FLUX.1 is a family of text-to-image generation models developed by Black Forest Labs, a German AI startup founded in 2024 by Robin Rombach, Andreas Blattmann, Patrick Esser, and Dominik Lorenz — the researchers who originally created Stable Diffusion at Stability AI.

What is the difference between FLUX.1 [schnell], [dev], and [pro]?

FLUX.1 [schnell] is the fastest model, fully open-source under Apache 2.0 for commercial and personal use. FLUX.1 [dev] is open-weight with 12B parameters but restricted to non-commercial use. FLUX.1 [pro] and FLUX1.1 [pro] are proprietary API models delivering the highest quality at $0.04–$0.06 per image.

How much does using FLUX via the API cost?

Black Forest Labs uses credit-based pricing where 1 credit = $0.01 USD. FLUX1.1 [pro] costs 4 credits ($0.04/image), Ultra/Raw modes cost 6 credits ($0.06/image), FLUX.1 Kontext [pro] costs 4 credits ($0.04/image), and FLUX.2 [klein] starts from $0.014/image with megapixel-based scaling.

Can I use FLUX models commercially?

Yes, FLUX.1 [schnell] is licensed under Apache 2.0 and can be used commercially for free. FLUX.2 [klein] models are also Apache 2.0. The [dev] models are for non-commercial use only. The Pro and API-based models are available commercially via the BFL API with per-image pricing.

Which major platforms and companies use FLUX?

FLUX powers image generation in Adobe Photoshop (Generative Fill uses FLUX.1 Kontext [pro]), Canva, Mistral AI's Le Chat, Vercel, ElevenLabs, VSCO, Freepik, Picsart, and was previously used by xAI's Grok chatbot. FLUX models are also available via fal.ai, Replicate, and Together AI.