Flux (text-to-image model)

Black Forest Labs was founded in 2024 by Robin Rombach, Andreas Blattmann, and Patrick Esser, former employees of Stability AI.{{Cite news |last=Killian |first=Nicolas |date=2024-08-27 |title=Black Forest Labs: Sie sind ein Teil von jener Kraft |url=https://www.zeit.de/digital/datenschutz/2024-08/black-forest-labs-ki-bildgenerator-elon-musk |url-access=subscription |url-status=live |archive-url=https://web.archive.org/web/20241004170543/https://www.zeit.de/digital/datenschutz/2024-08/black-forest-labs-ki-bildgenerator-elon-musk |archive-date=4 October 2024 |access-date=2024-11-17 |work=Die Zeit |language=de-DE |issn=0044-2070}}{{Cite web |last=Growcoot |first=Matt |date=2024-08-05 |title=AI Image Generator Made by Stable Diffusion Inventors on Par With Midjourney and DALL-E |url=https://petapixel.com/2024/08/05/ai-image-generator-made-by-stable-diffusion-inventors-on-par-with-midjourney-and-dall-e-flux1-black-forest-labs/ |access-date=2024-11-17 |website=PetaPixel |language=en}} All three founders had previously researched the artificial intelligence image generation at Ludwig Maximilian University of Munich as research assistants under Björn Ommer.{{Cite web |date=2024-08-22 |title=Black Forest Labs unter Beschuss: Schockierende KI-Bilder sorgen für… |url=https://www.alleaktien.com/news/black-forest-labs-unter-beschuss-schockierende-ki-bilder-sorgen-fur-emporung |access-date=2024-11-17 |website=AlleAktien |language=de}}{{Cite web |last=Hermes |first=Ann Kathrin |date=2024-08-08 |title=Black Forest Labs: KI-Tools aus dem Schwarzwald |url=https://www.trend.at/tools/black-forest-labs |access-date=2024-11-17 |website=trend.at |language=de}}{{Cite web |last=Schwär |first=Hannah |date=2024-08-15 |title=Black Forest Labs: Die Schwarzwald-KI, auf die Elon Musk setzt |url=https://www.capital.de/wirtschaft-politik/black-forest-labs--ki-aus-dem-schwarzwald-schlaegt-openai-34979066.html |access-date=2024-11-17 |website=Capital.de |language=de}} They published their research results on image generation in 2022, which resulted in creation of Stable Diffusion.{{Cite web |title=High-Resolution Image Synthesis with Latent Diffusion Models |url=https://ommer-lab.com/research/latent-diffusion-models/ |access-date=2024-11-17 |website=Computer Vision & Learning Group |language=en-US |archive-date=16 November 2024 |archive-url=https://web.archive.org/web/20241116192358/https://ommer-lab.com/research/latent-diffusion-models/ |url-status=live }} Investors in Black Forest Labs included venture capital firm Andreessen Horowitz, Brendan Iribe, Michael Ovitz, Garry Tan, and Vladlen Koltun. The company received an initial investment of {{Currency|31}}{{Nbsp}}million.{{Cite web |last=Steinschaden |first=Jakob |date=2024-08-12 |title=Black Forest Labs: 31 Mio. Dollar für Herausforderer von OpenAI und Midjourney |url=https://www.trendingtopics.eu/black-forest-labs-31-mio-dollar-fuer-herausforderer-von-openai-und-midjourney/ |access-date=2024-11-17 |website=Trending Topics |language=de |archive-date=28 August 2024 |archive-url=https://web.archive.org/web/20240828133326/https://www.trendingtopics.eu/black-forest-labs-31-mio-dollar-fuer-herausforderer-von-openai-und-midjourney/ |url-status=live }}{{Cite web |last=Nuñez |first=Michael |date=2024-08-01 |title=Stable Diffusion creators launch Black Forest Labs, secure $31M for FLUX.1 AI image generator |url=https://venturebeat.com/ai/stable-diffusion-creators-launch-black-forest-labs-secure-31m-for-flux-1-ai-image-generator/ |access-date=2024-11-17 |website=VentureBeat |language=en-US |archive-date=8 October 2024 |archive-url=https://web.archive.org/web/20241008120037/https://venturebeat.com/ai/stable-diffusion-creators-launch-black-forest-labs-secure-31m-for-flux-1-ai-image-generator/ |url-status=live }}

In August 2024, Flux was integrated into the Grok chatbot developed by xAI and made available as part of premium feature on X (formerly Twitter).{{Cite web |last=Puscher |first=Frank |title=Generative AI. Black Forest Labs und Flux.1: Vom Superstar zum Buhmann in fünf Tagen |url=https://meedia.de/news/beitrag/17472-black-forest-labs-und-flux-1-vom-superstar-zum-buhmann-in-fuenf-tagen.html |url-access=subscription |url-status=live |archive-url=https://web.archive.org/web/20240927210951/https://meedia.de/news/beitrag/17472-black-forest-labs-und-flux-1-vom-superstar-zum-buhmann-in-fuenf-tagen.html |archive-date=27 September 2024 |access-date=2024-11-17 |website=MEEDIA |language=de}}{{Cite web |last=Bomke |first=Luisa |title=Flux.1 – ein deutscher KI-Bildgenerator dreht mit Grok frei |url=https://www.handelsblatt.com/technik/ki/xai-kooperation-flux1-deutscher-ki-bildgenerator-dreht-mit-grok-frei/100059178.html |url-status=live |archive-url=https://web.archive.org/web/20240830103514/https://www.handelsblatt.com/technik/ki/xai-kooperation-flux1-deutscher-ki-bildgenerator-dreht-mit-grok-frei/100059178.html |archive-date=30 August 2024 |access-date=2024-11-17 |website=Handelsblatt |language=de}}{{Cite web |last=Weatherbed |first=Jess |date=2024-08-14 |title=xAI's new Grok-2 chatbots bring AI image generation to X |url=https://www.theverge.com/2024/8/14/24220127/grok-ai-chatbot-beta-image-generation-x-xai-update |url-status=live |archive-url=https://web.archive.org/web/20241117000048/https://www.theverge.com/2024/8/14/24220127/grok-ai-chatbot-beta-image-generation-x-xai-update |archive-date=17 November 2024 |access-date=2024-11-17 |website=The Verge |language=en}}{{Cite news |last=Metz |first=Rachel |date=2024-08-21 |title=This Tiny Startup Is Helping Musk's Grok With Image Generation |url=https://www.bloomberg.com/news/articles/2024-08-21/this-tiny-startup-is-helping-musk-s-grok-with-image-generation |url-access=subscription |access-date=2024-11-19 |work=Bloomberg News |language=en}} Grok later switched to its own text-to-image model Aurora in December 2024.{{Cite web |last=Davis |first=Wes |date=2024-12-07 |title=X gives Grok a new photorealistic AI image generator |url=https://www.theverge.com/2024/12/7/24315644/grok-x-aurora-ai-image-generator-xai |access-date=2024-12-10 |website=The Verge |language=en |archive-date=12 December 2024 |archive-url=https://web.archive.org/web/20241212193314/https://www.theverge.com/2024/12/7/24315644/grok-x-aurora-ai-image-generator-xai |url-status=live }}

On 18 November 2024, Mistral AI announced that its Le Chat chatbot had integrated Flux Pro as its image generation model.{{Cite web |date=2024-11-18 |title=Mistral has entered the chat |url=https://mistral.ai/news/mistral-chat/ |access-date=2024-12-11 |website=Mistral AI |language=en-us}}{{Cite web |last=Franzen |first=Carl |date=2024-11-18 |title=Mistral unleashes Pixtral Large and upgrades Le Chat into full-on ChatGPT competitor |url=https://venturebeat.com/ai/mistral-unleashes-pixtral-large-and-upgrades-le-chat-into-full-on-chatgpt-competitor/ |access-date=2024-12-11 |website=VentureBeat |language=en-US}}

On 21 November 2024, Black Forest Labs announced the release of Flux.1 Tools, a suite of editing tools designed to be used on top of existing Flux models. The tools consisting of Flux.1 Fill for inpainting and outpainting, Flux.1 Depth for control based on extracted depth map of input images and prompts, Flux.1 Canny for control based on extracted canny edges of input images and prompts, and Flux.1 Redux for mixing existing input images and prompts. Each tools are available in both Dev and Pro variants.{{Cite web |date=2024-11-21 |title=Introducing FLUX.1 Tools |url=https://blackforestlabs.ai/flux-1-tools/ |access-date=2024-12-13 |website=Black Forest Labs |language=en-US |archive-date=26 November 2024 |archive-url=https://web.archive.org/web/20241126040130/https://blackforestlabs.ai/flux-1-tools/ |url-status=live }}{{Cite web |last=Bastian |first=Matthias |date=2024-11-22 |title=Black Forest Labs expands FLUX.1 with four new AI tools for image editing |url=https://the-decoder.com/black-forest-labs-expands-flux-1-with-four-new-ai-tools-for-image-editing/ |access-date=2024-12-15 |website=The Decoder |language=en-US |archive-date=22 November 2024 |archive-url=https://web.archive.org/web/20241122224517/https://the-decoder.com/black-forest-labs-expands-flux-1-with-four-new-ai-tools-for-image-editing/ |url-status=live }}

In January 2025, Black Forest Labs announced a partnership with Nvidia for inclusion of Flux models as foundation models for Nvidia's Blackwell microarchitecture.{{Cite web |last=Takahashi |first=Dean |date=2025-01-07 |title=Nvidia unveils AI foundation models running on RTX AI PCs |url=https://venturebeat.com/ai/nvidia-unveils-ai-foundation-models-running-on-rtx-ai-pcs/ |access-date=2025-01-19 |website=VentureBeat |language=en-US |archive-date=9 January 2025 |archive-url=https://web.archive.org/web/20250109215759/https://venturebeat.com/ai/nvidia-unveils-ai-foundation-models-running-on-rtx-ai-pcs/ |url-status=live }} The company also announced the release of {{Proper name|Flux Pro Finetuning API}}, designed for customisation and fine-tuning of Flux-generated images and a partnership with German media company Hubert Burda Media for usage of Flux Pro as part of content creation.{{Cite web |last=Franzen |first=Carl |date=2025-01-16 |title=You can now fine-tune your own version of AI image maker Flux with just 5 images |url=https://venturebeat.com/ai/you-can-now-fine-tune-your-own-version-of-ai-image-maker-flux-with-just-5-images/ |access-date=2025-01-19 |website=VentureBeat |language=en-US |archive-date=24 January 2025 |archive-url=https://web.archive.org/web/20250124083654/https://venturebeat.com/ai/you-can-now-fine-tune-your-own-version-of-ai-image-maker-flux-with-just-5-images/ |url-status=live }}

On 29 May 2025, Black Forest Labs announced FLUX.1 Kontext, a suite of models that enable in-context image generation and editing, allowing users to prompt with both text and images.{{cite web|title=Introducing FLUX.1 Kontext and the BFL Playground|url=https://bfl.ai/announcements/flux-1-kontext|website=Black Forest Labs|date=29 May 2025 |access-date=|archive-date=|archive-url=}} Alongside this, they launched the BFL Playground, an interface for testing their FLUX models.

Flux is a series of text-to-image models. The models are based on rectified flow transformer blocks scaled to 12{{Nbsp}}billion parameters.{{Cite web |date=2024-08-01 |title=Announcing Black Forest Labs |url=https://blackforestlabs.ai/announcing-black-forest-labs/ |access-date=2024-11-17 |website=Black Forest Labs |language=en-US |archive-date=17 November 2024 |archive-url=https://web.archive.org/web/20241117160834/https://blackforestlabs.ai/announcing-black-forest-labs/ |url-status=live }} {{Cite web |date=2025-04-19 |title= black-forest-labs/FLUX.1-dev · Hugging Face |url=https://huggingface.co/black-forest-labs/FLUX.1-dev |access-date=2025-04-19 |website=Hugging Face |language=en-US}} The models are released under different licences with Schnell (meaning Fast or Quick in German language) released as open-source software under Apache License, Dev released as source-available software under a non-commercial licence, and Pro released as proprietary software and only available as API that can be licensed by third-party users.{{Cite web |title=Get Flux |url=https://blackforestlabs.ai/#get-flux |access-date=2024-11-17 |website=Black Forest Labs |language=en-US |archive-date=16 November 2024 |archive-url=https://web.archive.org/web/20241116053205/https://blackforestlabs.ai/#get-flux |url-status=live }}{{Cite web |last=Wiggers |first=Kyle |date=2024-10-03 |title=Black Forest Labs, the startup behind Grok's image generator, releases an API |url=https://techcrunch.com/2024/10/03/black-forest-labs-the-startup-behind-groks-image-generator-releases-an-api/ |access-date=2024-11-17 |website=TechCrunch |language=en-US |archive-date=4 October 2024 |archive-url=https://web.archive.org/web/20241004040824/https://techcrunch.com/2024/10/03/black-forest-labs-the-startup-behind-groks-image-generator-releases-an-api/ |url-status=live }} Users retained the ownership of resulting output regardless of models used.{{Cite web |title=flux/model_licenses/LICENSE-FLUX1-dev at main · black-forest-labs/flux |url=https://github.com/black-forest-labs/flux/blob/main/model_licenses/LICENSE-FLUX1-dev |access-date=2024-11-18 |website=GitHub |language=en |quote="Outputs. We claim no ownership rights in and to the Outputs. You are solely responsible for the Outputs you generate and their subsequent uses in accordance with this License. You may use Output for any purpose (including for commercial purposes), except as expressly prohibited herein. You may not use the Output to train, fine-tune or distill a model that is competitive with the FLUX.1 [dev] Model." |archive-date=15 September 2024 |archive-url=https://web.archive.org/web/20240915222202/https://github.com/black-forest-labs/flux/blob/main/model_licenses/LICENSE-FLUX1-dev |url-status=live }}{{Cite web |date=2024-08-01 |title=API Agreement - BFL Docs (Pro) |url=https://docs.bfl.ml/agreement/ |access-date=2024-11-18 |website=Black Forest Labs |quote=Output. Company claims no ownership rights in and to the Outputs, and Developer and Users may use the Output for their own personal or commercial purposes, subject to any restrictions set forth herein or in the Flux Service Terms. For the avoidance of doubt, Outputs do not include any components of the Flux API or the Flux AI model, such as its weights or parameters. |archive-date=3 October 2024 |archive-url=https://web.archive.org/web/20241003191037/https://docs.bfl.ml/agreement/ |url-status=live }}

The models can be used either online or locally by using generative AI user interfaces such as ComfyUI and Stable Diffusion WebUI Forge (a fork of Automatic1111 WebUI).{{cite web |author1=田口和裕 |date=18 August 2024 |title=話題の画像生成AI「FLUX.1」をStable Diffusion用の「WebUI Forge」で動かす（高速化も試してみました） (1/6) |url=https://ascii.jp/elem/000/004/215/4215695/ |access-date=21 November 2024 |website=ASCII.jp |publisher=ASCII Media Works |language=ja}}

An improved flagship model, Flux 1.1 Pro was released on 2 October 2024.{{cite web|title=Announcing FLUX1.1 [pro] and the BFL API|url=https://blackforestlabs.ai/announcing-flux-1-1-pro-and-the-bfl-api/|website=Black Forest Labs|date=2 October 2024 |access-date=2024-11-17|archive-date=|archive-url=}}{{Cite web |last=Franzen |first=Carl |date=2024-10-03 |title=Black Forest Labs releases Flux 1.1 Pro and an API |url=https://venturebeat.com/ai/black-forest-labs-releases-flux-1-1-pro-and-an-api/ |access-date=2024-11-17 |website=VentureBeat |language=en-US}} Two additional modes were added on 6 November, Ultra which can generate image at four times higher resolution and up to 4 megapixel without affecting generation speed and Raw which can generate hyper-realistic image in the style of candid photography.{{Cite web |last=Growcoot |first=Matt |date=2024-11-07 |title=Flux AI Introduces Raw Mode That 'Captures the Genuine Feel of Candid Photography' |url=https://petapixel.com/2024/11/07/flux-ai-introduces-raw-mode-that-captures-the-genuine-feel-of-candid-photography/ |access-date=2024-11-19 |website=PetaPixel |language=en}}{{Cite web |last=Bastian |first=Matthias |date=2024-11-06 |title=Flux 1.1 Pro AI image model adds "amateur" RAW photo mode and 4K image generation |url=https://the-decoder.com/flux-1-1-pro-ai-image-model-adds-amateur-raw-photo-mode-and-4k-image-generation/ |access-date=2024-11-17 |website=The Decoder |language=en-US}}{{Cite web |date=2024-11-06 |title=Introducing FLUX1.1 [pro] Ultra and Raw Modes |url=https://blackforestlabs.ai/flux-1-1-ultra/ |access-date=2024-11-17 |website=Black Forest Labs |language=en-US |archive-date=12 November 2024 |archive-url=https://web.archive.org/web/20241112145336/https://blackforestlabs.ai/flux-1-1-ultra/ |url-status=live }}

Related to Flux is text-to-video model SOTA, under development {{As of|2024|December|lc=y}}.

According to a test performed by Ars Technica, the outputs generated by Flux.1 Dev and Flux.1 Pro are comparable with DALL-E 3 in terms of prompt fidelity, with the photorealism closely matched Midjourney 6 and generated human hands with more consistency over previous models such as Stable Diffusion XL.

Flux has been criticised for its very realistic generated images. According to media reports, depictions ranged from an image of Donald Trump posing with guns to disturbing scenes, which triggered discussions about ethical implications of technologies developed by Black Forest Labs.

After the release of the model, social media platform X was flooded with Flux-generated images.{{Cite web |last=Zeff |first=Maxwell |date=2024-08-14 |title=Meet Black Forest Labs, the startup powering Elon Musk's unhinged AI image generator |url=https://techcrunch.com/2024/08/14/meet-black-forest-labs-the-startup-powering-elon-musks-unhinged-ai-image-generator/ |access-date=2024-11-17 |website=TechCrunch |language=en-US |archive-date=17 November 2024 |archive-url=https://web.archive.org/web/20241117003305/https://techcrunch.com/2024/08/14/meet-black-forest-labs-the-startup-powering-elon-musks-unhinged-ai-image-generator/ |url-status=live }}{{Cite web |last=Schwarzer |first=Matthias |date=2024-08-16 |title=Drogen, Bomben und Gewalt: KI-Bildgenerator von Elon Musk zeigt alles – mit deutscher Technik |url=https://www.rnd.de/digital/drogen-bomben-und-gewalt-ki-bildgenerator-von-elon-musk-zeigt-alles-mit-deutscher-technik-O4U7367ONNA4LF43GNMVNJARBI.html |access-date=2024-11-17 |website=RND.de |language=de}} Black Forest Labs has not provided exact details of the data used to train the model. Ars Technica suspected that Flux is based on a large, unauthorised collection of images scraped from the internet, a controversial practice with potential legal consequences.{{Cite web |last=Edwards |first=Benj |date=2024-08-02 |title=FLUX: This new AI image generator is eerily good at creating human hands |url=https://arstechnica.com/information-technology/2024/08/flux-this-new-ai-image-generator-is-eerily-good-at-creating-human-hands/ |access-date=2024-11-17 |website=Ars Technica |language=en-US}}{{Cite web |last=Künne |first=Christoph |date=2024-08-07 |title=FLUX.1: Neuer KI-Bildgenerator |url=https://www.docma.info/ki/flux1-neuer-ki-bildgenerator |access-date=2024-11-17 |website=DOCMA |language=de |archive-date=31 August 2024 |archive-url=https://web.archive.org/web/20240831113513/https://www.docma.info/ki/flux1-neuer-ki-bildgenerator |url-status=live }}

While Black Forest Labs do not offer direct access to their models on their website, the Flux models are widely available through various third-party platforms for creative and professional use. These include repositories on platforms like Hugging Face{{cite web | url=https://huggingface.co/black-forest-labs | title=Black Forest Labs on Hugging Face | website=Hugging Face | access-date=2025-01-05 | archive-date=2 February 2025 | archive-url=https://web.archive.org/web/20250202084151/https://huggingface.co/black-forest-labs | url-status=live }} and Replicate.{{cite web | url=https://replicate.com/collections/flux | title=Flux Collection on Replicate | website=Replicate | access-date=2025-01-05 | archive-date=25 January 2025 | archive-url=https://web.archive.org/web/20250125142958/https://replicate.com/collections/flux | url-status=live }}

{{Commons category}}

• {{Official website|https://bfl.ai/}}
• [https://huggingface.co/black-forest-labs Flux models on Hugging Face]
• [https://replicate.com/collections/flux Flux models on Replicate]
• [https://fal.ai/models?keywords=flux Flux models on FAL.ai]

{{Generative AI}}

{{Artificial intelligence navbox}}

Category:Unsupervised learning

Category:Text-to-image generation

Category:Deep learning software applications

Category:Artificial intelligence art

Category:Open-source artificial intelligence

Category:2024 establishments in Germany

Category:Software using the Apache license