| Flux | |
|---|---|
.webp) An image generated with Flux.1 Kontext Pro. Partial prompt: A tranquil, sunlit mountain landscape in the style of a digital painting | |
| Original author | Black Forest Labs |
| Developer | Black Forest Labs |
| Initial release | August 2024 |
| Stable release | Flux.1 Kontext (model series)[1] / 29 May 2025 |
| Repository | |
| Type | Text-to-image model |
| License |
|
| Website | bfl |
Flux (also known asFLUX.1) is atext-to-image model developed by Black Forest Labs (BFL), based inFreiburg im Breisgau, Germany. Black Forest Labs was founded by former employees ofStability AI. As with other text-to-image models, Fluxgeneratesimages fromnatural language descriptions, calledprompts.
Black Forest Labs (BFL) was founded in 2024 by Robin Rombach, Andreas Blattmann, and Patrick Esser, former employees of Stability AI.[2][3] All three founders had previously researched the artificial intelligence image generation atLudwig Maximilian University of Munich as research assistants under Björn Ommer.[4][5][6] They published their research results on image generation in 2022, which resulted in creation ofStable Diffusion.[6][7] Investors in BFL included venture capital firmAndreessen Horowitz,Brendan Iribe,Michael Ovitz,Garry Tan, andVladlen Koltun.[8] The company received an initial investment ofUS$31 million.[9][10]
In August 2024, Flux was integrated into theGrok chatbot developed byxAI and made available as part of premium feature onX (formerly Twitter).[11][12][13][14] Grok later switched to its own text-to-image modelAurora in December 2024.[15]
On 18 November 2024,Mistral AI announced that itsLe Chat chatbot had integrated Flux Pro as its image generation model.[16][17]
On 21 November 2024, BFL 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 forinpainting and outpainting, Flux.1 Depth for control based on extracteddepth map of input images and prompts, Flux.1 Canny for control based on extractedcanny edges of input images and prompts, and Flux.1 Redux for mixing existing input images and prompts. Each tools are available in bothPro andDev models.[18][19]
In January 2025, BFL announced a partnership withNvidia for inclusion of Flux models as foundation models for Nvidia'sBlackwell microarchitecture.[20] The company also announced the release of Flux Pro Finetuning API, designed for customisation andfine-tuning of Flux-generated images and a partnership with German media companyHubert Burda Media for usage of Flux Pro as part of content creation.[21]
On 29 May 2025, BFL 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.[22][23] Alongside this, BFL Playground, an interface for testing Flux models was released.[22][23]
On 31 July 2025, BFL announced Flux.1 Krea Dev, a model developed in collaboration with Krea AI that trained to achieve better performance, more varied aesthetics, and better realism compared to existing text-to-image models.[24]
In September 2025,Adobe Inc. announced thatPhotoshop (beta) users can use Flux.1 Kontext Pro as a model for its generative fill tool.[25] BFL collaborated withMeta on Vibes, anAI slop video-generation app.[26]
.webp)
.webp)
Flux is a series of text-to-image models. The models are based on rectified flow transformer blocks scaled to 12 billion parameters.[8][27] Flux.1 models were released under different licences withSchnell (meaning Fast or Quick inGerman language) released asopen-source software underApache License,Dev released assource-available software under a non-commercial licence (users can obtain a self-serving commercial licence forDev from BFL), andPro released asproprietary software and only available asAPI that can be licensed by third-party users.[28][29] Users retained the ownership of resulting output regardless of models used.[30][31]
The models can be used either online or locally by using generative AI user interfaces such asComfyUI andStable Diffusion WebUI Forge (afork of Automatic1111 WebUI).[8][32]
An improved flagship model, Flux 1.1 Pro was released on 2 October 2024.[33][34] 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 ofcandid photography.[35][36][37]
Flux.1 Kontext is a series with in-context image generation and editing capabilities. It is available inMax,Pro, andDev models.Max is the highest quality model and can be used to iteratively modify an existing image by using prompt whilePro is optimized to balance quality and speed of generation.[22]Dev is an open-weight model released under non-commercial license, same as Flux.1 Dev.[38]
Related to Flux istext-to-video model SOTA, under development as of June 2025[update].[8]
According to a test performed byArs Technica, the outputs generated by Flux.1 Dev and Flux.1 Pro are comparable withDALL-E 3 in terms of prompt fidelity, with the photorealism closely matchedMidjourney 6 and generated human hands with more consistency over previous models such as Stable Diffusion XL.[39]
Flux has been criticised for its very realistic generated images. According to media reports, depictions ranged from an image ofDonald Trump posing with guns to disturbing scenes, which triggered discussions about ethical implications of Flux models.[4][13]
After the release of the model, social media platformX was flooded with Flux-generated images.[40][41] Black Forest Labs has not provided exact details of the data used to train the model.[35]Ars Technica suspected that Flux is based on a large, unauthorised collection of imagesscraped from the internet, a controversial practice with potential legal consequences.[39][42]
According to a test performed by Japanese technology news websiteGigazine for Flux.1 Kontext, the model series has a good understanding of the English language and can easily transferstyle of the image from photorealistic into anime-style according to prompts given by the user; however, its ability to understandJapanese is quite poor.[43]
In addition to the official BFL Playground on its website,[44] the Flux models are also widely available through various third-party platforms for creative and professional use. These include repositories on platforms likeHugging Face[45] and Replicate.[46]
1a. "Derivative" means any (i) modified version of the FLUX.1 [dev] Model (including but not limited to any customized or fine-tuned version thereof), (ii) work based on the FLUX.1 [dev] Model, or (iii) any other derivative work thereof. For the avoidance of doubt, Outputs are not considered Derivatives under this License." "2d. 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 or the FLUX.1 Kontext [dev] Model.
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.
| Concepts |
| ||||||
|---|---|---|---|---|---|---|---|
| Applications | |||||||
| Implementations |
| ||||||
| People |
| ||||||
| Architectures |
| ||||||
