正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Introducing FLUX 3. One multi-modal model for Image, Video, Audio and Action-Prediction. Creations are truer to life in every kind of style. FLUX 3 Video is now available in early access (link below). Jointly trained in one unified architecture, our model can be extended to predict actions for robotics.... show more

Black Forest Labs

56,513 subscribers

1,219,591 次观看 • 3 个月前 •via X (Twitter)

科学技术教育新闻政治

Anya Rossi• Live Now

Private livecam show

38 条评论

Black Forest Labs 的头像

Black Forest Labs3 个月前

Request Early Access to FLUX 3 Video:

Black Forest Labs 的头像

Black Forest Labs3 个月前

Over the next few weeks and months, we’ll make the following capabilities available, each after an early access phase for ensuring smooth rollout: • Video with native audio generation (now in early access). • Action prediction through selected research and commercial partners. Beginning with mimic robotics. • Image generation and editing (sneak peek below 🎄). • Fast variants and features for iterating cost efficiently. • Open-weight access to a multimodal backbone, for content creation (video, audio and image) and action prediction. FLUX 3 is a checkpoint on our mission to develop real-world visual intelligence: models that perceive, predict, and act across digital and physical environments. Learn more about FLUX 3:

Black Forest Labs 的头像

Black Forest Labs3 个月前

Action: An early version of FLUX 3 is now running on robots. @mimicrobotics was one of the first partners to gain early access to FLUX 3. Together we developed FLUX-mimic, a video-action model combining the FLUX 3 backbone with mimic's expertise in robot learning for dexterous manipulation and production deployment. It’s now running robots that have been tested and deployed at Audi. Read our thesis on why physical AI and content creation run on the same foundation:

Black Forest Labs 的头像

Black Forest Labs3 个月前

P.S. some early samples from FLUX 3 Image 👀

Christopher Fryant 的头像

Christopher Fryant3 个月前

Thanks for giving me the chance to try it out in early access, it's pretty great! Got me staying up til 2 AM trying everything I could. Hasn't happened in a long time!

Osho 的头像

Osho3 个月前

this a few hours ago!

Troy 的头像

Troy3 个月前

FLUX is great for fashion photography. Can't wait to try FLUX 3!

Stephen 的头像

Stephen3 个月前

Only 2 years delayed, soon™️

Ed Newton-Rex 的头像

Ed Newton-Rex3 个月前

What’s the training data?

Simul S. 的头像

Simul S.3 个月前

One of the best AI generated video, I ever seen, freaking beautiful.

Chris B. Ward (e/bored) 🇬🇧🇮🇪🤐 的头像

Chris B. Ward (e/bored) 🇬🇧🇮🇪🤐3 个月前

1. Can it be run locally? 2. If so, how much VRAM is required?

Brian Rayaguna 的头像

Brian Rayaguna3 个月前

@bfl_ai :

Filecoin 的头像

Filecoin3 个月前

action prediction for robotics produces deployment data that must remain verifiable and durable so the model can improve over time.

will depue 的头像

will depue3 个月前

the videos look soo good. amazing work, especially on post-training here

Dreaming Tulpa 🥓👑 的头像

Dreaming Tulpa 🥓👑3 个月前

oh yeah, would love to test this for making my next music video!

Joe Muller 的头像

Joe Muller3 个月前

Open source it

Matt White 的头像

Matt White3 个月前

What the Flux @bfl_ai? This is amazing! Can't wait! Two questions: 1. Will Flux 3 be open source? 2. Will you use a permissive model license (ie. OpenMDW)?

Alex Patrascu 的头像

Alex Patrascu3 个月前

Here. We. Go!

ai lover fun wan 的头像

ai lover fun wan3 个月前

Please release a small model 14b / 22b open source in comfyui to compete with ltx 2.3 and wan models.

Ivan Fioravanti ᯅ 的头像

Ivan Fioravanti ᯅ3 个月前

This is more than amazing!

kzn 的头像

kzn3 个月前

not open source? don't care.

The Dark Developer 的头像

The Dark Developer3 个月前

More Examples, 11 minutes of Flux 3 videos

Wetterschneider 的头像

Wetterschneider3 个月前

Running horses? Sharks underwater? Waves splashing on a cliff? Monk lighting candles? Glass shattering? Office chair racing in parking garage? Science? Snowboarder? Almost every frame is a copy of existing footage that this software was trained on. There was a whale in the desert so that was fine. Congrats.

Raiven 的头像

Raiven3 个月前

Any plans for a dev model?

Michael Swengel 的头像

Michael Swengel3 个月前

I wonder how many of these replies are from bots.

Elvis Nava 的头像

Elvis Nava3 个月前

LFG!

Eth 的头像

Eth3 个月前

Considering Flux.2 klein is still one of the best models out there - specially in prompt adherence - for us local nuts, I really hope you guys can pull a small OS version that we can play with.

kriGeta 的头像

kriGeta3 个月前

Yo @ostrisai I hope you are cooking lora training…🔥

Umesh 的头像

Umesh3 个月前

Finally! Congratulations 👏

J A Z I I 的头像

J A Z I I3 个月前

hi ser can you lemme in please?

Heather Cooper 的头像

Heather Cooper3 个月前

I've been waiting for this for a couple of years😊

Shikhar 的头像

Shikhar3 个月前

world models are coming

kiri49 的头像

kiri493 个月前

future open weights? better than what @Alibaba_Qwen is doing which is no open weights for qwen image 3 at all

Prashanth 的头像

Prashanth3 个月前

AI figured out hands and immediately got a factory job

Stephen 的头像

Stephen3 个月前

@IamEmily2050 it's here!

Haunted AI 的头像

Haunted AI3 个月前

Yes! Let go!

lost in latency 的头像

lost in latency3 个月前

im in love, congrats on the release really excited to try!

PolyMe 的头像

PolyMe3 个月前

Congrats from a team building on FLUX. Schnell has been the image stage of our local text to 3D pipeline for months, the reference image quality decides the whole mesh downstream. Very curious how 3 handles single subject full body shots.

相关视频

FLUX 3 Video is live in ComfyUI via Partner Nodes Black Forest Labs FLUX 3 Model highlights: → 20s in a single generation → native audio, not dubbed on after → text-to-video, image-to-video, keyframes, continuation → multi-scene and multi-angle in one prompt → multilingual dialogue with lipsync that holds → typography that sits in the scene → chain clips agentically for longer stories One model across video, audio, image, and action. Click the link below to download the workflows and learn more with blog

FLUX 3 Video is live in ComfyUI via Partner Nodes Black Forest Labs FLUX 3 Model highlights: → 20s in a single generation → native audio, not dubbed on after → text-to-video, image-to-video, keyframes, continuation → multi-scene and multi-angle in one prompt → multilingual dialogue with lipsync that holds → typography that sits in the scene → chain clips agentically for longer stories One model across video, audio, image, and action. Click the link below to download the workflows and learn more with blog

ComfyUI

26,532 次观看 • 3 个月前

Introducing FLUX-mimic, a next-generation Video-Action Model for general purpose dexterity, developed in partnership with Black Forest Labs. Late last year we published mimic-video and introduced Video-Action Models (VAM): a new family of robotics foundation models built on top of video generation models. We showed that robot control reduces to visual prediction, and that robot capability is downstream of improvements in video modeling accuracy. The obvious implication was that advances in the video modeling frontier would directly translate to increased capabilities in end-to-end robot learning. FLUX-mimic is that thesis at frontier scale: We've applied our VAM architecture to the strongest video backbone available today, FLUX 3 from Black Forest Labs, and trained it on data from our own robots and wearables. General-purpose dexterity, running on a single GPU on premises. Because the model already understands world dynamics, it needs far fewer demonstrations to learn a new task. This is game-changing for our mission to deploy robots to factory floors, where industrial robot data is scarce and expensive to collect. We're now testing and deploying FLUX-mimic with manufacturing leaders like Audi USA, on complex, multi-step manipulation long considered impossible for conventional automation.

Introducing FLUX-mimic, a next-generation Video-Action Model for general purpose dexterity, developed in partnership with Black Forest Labs. Late last year we published mimic-video and introduced Video-Action Models (VAM): a new family of robotics foundation models built on top of video generation models. We showed that robot control reduces to visual prediction, and that robot capability is downstream of improvements in video modeling accuracy. The obvious implication was that advances in the video modeling frontier would directly translate to increased capabilities in end-to-end robot learning. FLUX-mimic is that thesis at frontier scale: We've applied our VAM architecture to the strongest video backbone available today, FLUX 3 from Black Forest Labs, and trained it on data from our own robots and wearables. General-purpose dexterity, running on a single GPU on premises. Because the model already understands world dynamics, it needs far fewer demonstrations to learn a new task. This is game-changing for our mission to deploy robots to factory floors, where industrial robot data is scarce and expensive to collect. We're now testing and deploying FLUX-mimic with manufacturing leaders like Audi USA, on complex, multi-step manipulation long considered impossible for conventional automation.

mimic

119,810 次观看 • 3 个月前

Introducing FLUX 3 Action. An open weights 7B World Action Model that achieves first place on the RoboLab benchmark. It outperforms the previous best open model by 6.1 percentage points while using 56% fewer parameters and running up to 3.95x faster.⁠⁠ FLUX 3 Action removes the usual trade-off between world action model performance and VLA speed: it still predicts video and actions together, but plans more than twice as far ahead and runs faster per second of robot motion than the strongest open VLA. Teams can fine-tune FLUX 3 Action on their own demonstrations to create policies for a particular robot and task. Together with NVIDIA, we also integrated FLUX 3 Action natively into Hugging Face's LeRobot, with fine-tuning recipes included and edge deployment on NVIDIA Jetson. Beyond robotics, we’re also seeing promising results training task-specific policies for acting in simulated environments like gaming, controlling a vehicle, computer use, and wherever else a model needs to understand a visual environment and then choose what to do next. FLUX 3 Action builds on the same image, video, and audio pretraining as FLUX 3, but uses a smaller architecture designed for practical deployment. In midtraining, we trained the model to predict actions and future frames together. We’re releasing the weights, code, fine-tuning recipe, benchmarks, and reproducible examples so researchers and developers can build on the model with their own robots, environments, and tasks (see below).

Introducing FLUX 3 Action. An open weights 7B World Action Model that achieves first place on the RoboLab benchmark. It outperforms the previous best open model by 6.1 percentage points while using 56% fewer parameters and running up to 3.95x faster.⁠⁠ FLUX 3 Action removes the usual trade-off between world action model performance and VLA speed: it still predicts video and actions together, but plans more than twice as far ahead and runs faster per second of robot motion than the strongest open VLA. Teams can fine-tune FLUX 3 Action on their own demonstrations to create policies for a particular robot and task. Together with NVIDIA, we also integrated FLUX 3 Action natively into Hugging Face's LeRobot, with fine-tuning recipes included and edge deployment on NVIDIA Jetson. Beyond robotics, we’re also seeing promising results training task-specific policies for acting in simulated environments like gaming, controlling a vehicle, computer use, and wherever else a model needs to understand a visual environment and then choose what to do next. FLUX 3 Action builds on the same image, video, and audio pretraining as FLUX 3, but uses a smaller architecture designed for practical deployment. In midtraining, we trained the model to predict actions and future frames together. We’re releasing the weights, code, fine-tuning recipe, benchmarks, and reproducible examples so researchers and developers can build on the model with their own robots, environments, and tasks (see below).

Black Forest Labs

182,671 次观看 • 1 个月前

Black Forest Labs just announced FLUX 3: a unified multimodal model for image, video, audio and action prediction. Founded in Freiburg, Germany, one of the few globally significant European companies in the AI sector. The current release is gated early access, not open source. But if FLUX 3 Dev actually ships with usable open weights, this could become one of the most important open-model releases in generative video yet. Especially after the Chinese Minimax H3 release. BFL claims, based on preliminary internal comparisons, that FLUX 3 outperformed models such as Runway Gen-4.5, Luma Ray 3.2, Kling 3 Pro, and Seedance 2.0. However, these figures stem from the early-access phase and do not constitute independent validation. Be that as it may, it's great to see the quality that today's video models can produce at affordable prices. Something that would have been unthinkable a year ago. Really cool release!

Black Forest Labs just announced FLUX 3: a unified multimodal model for image, video, audio and action prediction. Founded in Freiburg, Germany, one of the few globally significant European companies in the AI sector. The current release is gated early access, not open source. But if FLUX 3 Dev actually ships with usable open weights, this could become one of the most important open-model releases in generative video yet. Especially after the Chinese Minimax H3 release. BFL claims, based on preliminary internal comparisons, that FLUX 3 outperformed models such as Runway Gen-4.5, Luma Ray 3.2, Kling 3 Pro, and Seedance 2.0. However, these figures stem from the early-access phase and do not constitute independent validation. Be that as it may, it's great to see the quality that today's video models can produce at affordable prices. Something that would have been unthinkable a year ago. Really cool release!

Chubby♨️

39,869 次观看 • 3 个月前

Two years later our first video model - FLUX 3 Video - is finally released🥹. This is such a strong, fun and versatile model made by the best team in the world Black Forest Labs. We will add further variants and generation from references in the near future. I am also super excited about the upcoming open weight, action prediction and image variants of FLUX 3.

Two years later our first video model - FLUX 3 Video - is finally released🥹. This is such a strong, fun and versatile model made by the best team in the world Black Forest Labs. We will add further variants and generation from references in the near future. I am also super excited about the upcoming open weight, action prediction and image variants of FLUX 3.

Robin Rombach

26,359 次观看 • 3 个月前

The reception since we launched the first unified audio + text + image to video model in production last week has been incredible! We wanted to take a moment and highlight some firsts: - the first omnimodal modal that jointly operates over text, image, and audio. Fun fact: we shipped our first audio-to-video model Character-1 last June! - Character-3 is generating 5 minute uncut clips, and capable of temporally stable longform prediction. - Character-3 handles realistic humans, cartoons, side profiles, dynamic backgrounds, anime, animals, and more. - Character-3 sets a new standard for fast & efficient video diffusion at HD and SD resolutions, leveraging our advancements in realtime video generation. We'll have an announcement soon, but you can read more about an earlier version of this line of work in our paper here:

The reception since we launched the first unified audio + text + image to video model in production last week has been incredible! We wanted to take a moment and highlight some firsts: - the first omnimodal modal that jointly operates over text, image, and audio. Fun fact: we shipped our first audio-to-video model Character-1 last June! - Character-3 is generating 5 minute uncut clips, and capable of temporally stable longform prediction. - Character-3 handles realistic humans, cartoons, side profiles, dynamic backgrounds, anime, animals, and more. - Character-3 sets a new standard for fast & efficient video diffusion at HD and SD resolutions, leveraging our advancements in realtime video generation. We'll have an announcement soon, but you can read more about an earlier version of this line of work in our paper here:

Hedra

180,695 次观看 • 1 年前

FLUX 3 Video is #2 in the world. To celebrate, FLUX 3 Video is free to use in our playground until Sunday 16th, 11:59pm PT (link in the thread). This is just the beginning of what’s coming. Up next: 4K, video editing, and multiple images & videos as reference input.

FLUX 3 Video is #2 in the world. To celebrate, FLUX 3 Video is free to use in our playground until Sunday 16th, 11:59pm PT (link in the thread). This is just the beginning of what’s coming. Up next: 4K, video editing, and multiple images & videos as reference input.

Black Forest Labs

410,751 次观看 • 3 个月前

FLUX Video Upscale: 2K and 4K for Video Create videos in 2K or 4K with sharper faces, cleaner textures, and finer background detail. FLUX 3 Video’s native upscaling is designed to retain FLUX 3’s style diversity, natural motion, and realism at higher resolutions, with faster results than 3rd party upscalers. You can also use it to upscale any video. Link to use it in the thread

FLUX Video Upscale: 2K and 4K for Video Create videos in 2K or 4K with sharper faces, cleaner textures, and finer background detail. FLUX 3 Video’s native upscaling is designed to retain FLUX 3’s style diversity, natural motion, and realism at higher resolutions, with faster results than 3rd party upscalers. You can also use it to upscale any video. Link to use it in the thread

Black Forest Labs

50,948 次观看 • 2 个月前

Today, we are excited to release FLUX.1 Tools, a suite of models designed to add control and steerability to our base text-to-image model FLUX.1, enabling the modification and re-creation of real and generated images. Learn more in our blogpost:

Today, we are excited to release FLUX.1 Tools, a suite of models designed to add control and steerability to our base text-to-image model FLUX.1, enabling the modification and re-creation of real and generated images. Learn more in our blogpost:

Black Forest Labs

350,563 次观看 • 1 年前

Flux 3 Max is now available in ComfyUI Generate a 10-second, 720/768p video in ~12 seconds. Click the link below to run the workflows ⬇️

Flux 3 Max is now available in ComfyUI Generate a 10-second, 720/768p video in ~12 seconds. Click the link below to run the workflows ⬇️

ComfyUI

29,998 次观看 • 20 天前

Meet Higgsfield × FLUX 3 Image. With Black Forest Labs' new image model, you can: > edit one detail without touching the rest > place objects with bounding boxes > combine up to 10 references > generate in up to 4K Available now on Higgsfield.

Meet Higgsfield × FLUX 3 Image. With Black Forest Labs' new image model, you can: > edit one detail without touching the rest > place objects with bounding boxes > combine up to 10 references > generate in up to 4K Available now on Higgsfield.

Higgsfield AI 🧩

85,293 次观看 • 1 个月前

FLUX Premium is now available in LTX Studio. This new image model brings major improvements to visual quality, style control, and prompt accuracy. Exclusively for Standard, Pro, and Enterprise users. Try it out today.

FLUX Premium is now available in LTX Studio. This new image model brings major improvements to visual quality, style control, and prompt accuracy. Exclusively for Standard, Pro, and Enterprise users. Try it out today.

LTX.io

1,031,013 次观看 • 1 年前

kinda in love with flux 3 the model has a similar feeling to grok imagine 0.9 which i absolutely adored can't wait for the open weights and hopefully more unrestricted access shaping up to be a banger this one

kinda in love with flux 3 the model has a similar feeling to grok imagine 0.9 which i absolutely adored can't wait for the open weights and hopefully more unrestricted access shaping up to be a banger this one

Dreaming Tulpa 🥓👑

17,595 次观看 • 3 个月前

The first truly open-source audio-video model. LTX-2 is a DiT-based foundation model with all core video generation capabilities in one unified model. Designed to run locally on consumer GPUs. - text-to-video - image-to-video - and video-to-video modes 100% open-source.

The first truly open-source audio-video model. LTX-2 is a DiT-based foundation model with all core video generation capabilities in one unified model. Designed to run locally on consumer GPUs. - text-to-video - image-to-video - and video-to-video modes 100% open-source.

Akshay 🚀

66,268 次观看 • 10 个月前

FLUX 3 Max is now available on Built for speed: · 10-second videos in about 12 seconds · Text-to-video and image-to-video · Fast enough to iterate on every shot Less time waiting on renders, more time choosing the best take.

FLUX 3 Max is now available on Built for speed: · 10-second videos in about 12 seconds · Text-to-video and image-to-video · Fast enough to iterate on every shot Less time waiting on renders, more time choosing the best take.

Sunra

20,141 次观看 • 20 天前

FLUX.2 [klein] 4B + 9B are the newest BFL models in the Flux family—combining image generation + image editing in one compact architecture. Built for interactive workflows and quick iteration, with extremely fast inference on distilled variants. Two models. Two modes. ⚡️

FLUX.2 [klein] 4B + 9B are the newest BFL models in the Flux family—combining image generation + image editing in one compact architecture. Built for interactive workflows and quick iteration, with extremely fast inference on distilled variants. Two models. Two modes. ⚡️

ComfyUI

19,899 次观看 • 9 个月前

FLUX 3 is now available on Runway. Generate and edit up to 20 seconds of video with audio. Try it today.

FLUX 3 is now available on Runway. Generate and edit up to 20 seconds of video with audio. Try it today.

Runway

35,498 次观看 • 3 个月前

Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more. Now available for free on all our products:

Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more. Now available for free on all our products:

SpaceX Super Intelligence

8,319,073 次观看 • 1 年前

Introducing Ideogram Character -- the first character consistency model that works with just one reference image. Now available to all users for free! Create consistent character images in any style, expression, scene, and lighting. Who wants video next? 🥺

Introducing Ideogram Character -- the first character consistency model that works with just one reference image. Now available to all users for free! Create consistent character images in any style, expression, scene, and lighting. Who wants video next? 🥺

Ideogram

1,714,484 次观看 • 1 年前

Today, we are releasing Stable Video Diffusion, our first foundation model for generative AI video based on the image model, Stable Diffusion. As part of this research preview, the code, weights, and research paper are now available. Additionally, today you can sign up for our waitlist to access a new upcoming web experience featuring a Text-To-Video interface. To access the model & sign up for our waitlist, visit our website here:

Today, we are releasing Stable Video Diffusion, our first foundation model for generative AI video based on the image model, Stable Diffusion. As part of this research preview, the code, weights, and research paper are now available. Additionally, today you can sign up for our waitlist to access a new upcoming web experience featuring a Text-To-Video interface. To access the model & sign up for our waitlist, visit our website here:

Stability AI

1,024,917 次观看 • 3 年前