Black Forest Labs Unveils FLUX 3, A New Multimodal Frontier Model For Visual Intelligence

Jointly trained across image, video, audio, and action prediction modalities within a unified architecture, FLUX 3 brings a coherent understanding of the real world to every kind of visual creation—from generative media to robotics and beyond

FREIBURG, Germany, July 23, 2026 (GLOBE NEWSWIRE) -- Black Forest Labs, the global frontier AI research lab building the foundation layer for visual intelligence, today introduces FLUX 3, its new multimodal frontier model. FLUX 3 jointly learns from images, video, and audio within a unified architecture, and can also be extended to predict actions. The model is Black Forest Labs’ newest addition to its FLUX family of visual AI models, which are known for pairing state-of-the-art capabilities with open access, and for powering generative features inside leading creative, developer, and consumer platforms like Adobe Photoshop, Picsart, Nous Research’s Hermes Agent, and more.

The unveiling of FLUX 3 comes at an inflection point for the industry, as the focus moves beyond language models toward intelligent systems that can understand both the visual and physical world. While this shift has produced a cluster of adjacent fields—image and video generation, world models, simulation, robotics and physical AI—Black Forest Labs sees these categories as interconnected expressions of the same underlying capability: visual intelligence, or models that can perceive, predict, and act across physical and digital environments.

Multimodality for Perception and Simulation
The world is multimodal. Intelligent species and systems have evolved with the capability to process many kinds of information, including vision, sound, and other sensory signals. Vision in particular is the human brain's highest-bandwidth input system. But perception is only one component of intelligence; equally as important is the ability to form a coherent representation of the world, and to act within it.

“We place vision at the center of our approach because it is the most signal-rich medium of the physical world. Images convey structure, images and video teach spatial relationships, video teaches dynamics, and actions reveal causal relationships. But vision alone is not the complete picture,” said Robin Rombach, Co-Founder and CEO of Black Forest Labs. “True intelligence means perceiving the world: predicting how it will change, taking action, and learning from the results. Joint training within one unified architecture is what will get us there, because each training modality strengthens the others. Audio conveys timing, prosody, and physical events that elude vision. Language conveys goals, abstractions, and instructions that pixels cannot easily express.”

One Model, Multiple Capabilities
FLUX 3 builds on Self-Flow, the company’s breakthrough approach for efficiently aligning multimodal generation and understanding within the same underlying architecture. By scaling up compute and data resources, FLUX 3 was trained across multiple modalities at the same time. Testing ultimately showed that generative video generation and action prediction do not actually require separate foundations, as the same underlying architecture could be extended to action prediction without sacrificing the capabilities learned from videos.

“You can’t cheat reality. A model that only learns images can only generate images. But the world is not made of still frames. It moves, sounds, changes, and responds,” said Rombach. “That’s why FLUX 3 is trained across those signals together, so the model can build a deeper understanding of how the world works. That is what visual intelligence requires, whether the application is video generation, simulation, or robotics.”

FLUX 3 is built to support applications including in creative tooling, media, design, e-commerce, and physical AI—including video generation with synchronized audio, precisely edited images, the maintenance of product and material consistency across motion, and action prediction for robotics.

The model will be available through FLUX 3 Video, FLUX 3 Image, FLUX 3 Action, and FLUX 3 Dev. FLUX 3 is already being tested by Canva, Burda, Magnific (formerly Freepik), Krea, and Picsart. While still in development, FLUX 3 Video already leads in early evaluations against frontier video models, and is particularly strong in capturing human facial expressions, associating sounds with physical events, and multilingual capabilities.

Visual Intelligence in the Real World
Black Forest Labs and mimic robotics today also introduced FLUX-mimic, a video-action model built on FLUX 3. FLUX-mimic is currently undergoing testing and deployment with leading manufacturing companies, like Audi.

FLUX-mimic is designed for general-purpose robotic manipulation: helping robots understand a visual scene, predict the consequences of an action, and adapt to new tasks with far less task-specific data. Depending on task difficulty, the model can be fine-tuned for a specific manipulation task with as little as 30 minutes of robot data, where prior approaches have required 30 or more hours.

“The hardest part of robotics is data,” said Elvis Nava, CTO of mimic robotics. “Every new task normally means hours of a robot repeating itself. Because FLUX-mimic is built on top of frontier video models that already understand how the physical world behaves, it picks up a new task in minutes, not days. This way, we can leapfrog the current state of the art in robot learning."

“In partnership with mimic, Audi has been testing and deploying FLUX-mimic. We have seen these robots solve complex soft body manipulation work that would have been simply impossible with conventional robotics. This can have a major impact in assisting our employees, increasing efficiency, and expanding flexible automation across production and logistics operations,” said Christoph Schneider, Audi Production Lab. “For us, partnering with pioneering companies such as mimic and Black Forest Labs is essential in pushing the frontier of Physical AI and validating these innovations in real-world production environments.”

FLUX 3 Availability
FLUX 3 Video with optional native audio generation and FLUX 3 Action are now available for early access, while FLUX 3 Image will roll out in the coming weeks. Developers and teams can apply for access here, with general availability to follow. Black Forest Labs will publish full FLUX 3 benchmark results and methodology alongside broader availability.

Black Forest Labs will also release faster and open-weight versions of FLUX 3 later this year, consistent with the company’s long-held belief in building and sharing its technology openly so researchers and developers around the world can build, test, and deploy new applications on top of it. Open weights make secure, low-latency local deployment possible for applications like robotic control systems, and let teams adapt FLUX 3 to their own data, products, and workflows—turning a general model into a secure foundation for any creative, simulation, or physical AI stack.

“We are only beginning to scratch the surface of versatile, capable, unified visual models,” said Rombach. “From interactive image and video editing to simulation, physical AI, and computer use, the frontier is wide open.”

Black Forest Labs has quickly become a standard visual AI engine across the builder and creative ecosystems since its 2024 launch from stealth. Its founders have been pioneering modern visual AI for far longer, from VQGAN, latent diffusion, and Stable Diffusion to Black Forest Labs’ family of FLUX models, which power AI capabilities for leading enterprise platforms and creative industry professionals like legendary film director Martin Scorsese. To date, models from the Black Forest Labs founders have been downloaded over half a billion times.

About Black Forest Labs
Black Forest Labs is a global frontier AI research lab building the foundation layer for visual intelligence. Founded by the researchers who pioneered latent diffusion, Stable Diffusion and the FLUX model family, the 100-person team advances research and ships open, production-ready models from Freiburg, Germany and San Francisco, USA. Valued at $3.25 billion, the company has raised more than $450 million from leading venture and strategic investors including a16z, AMP, Salesforce Ventures, NVIDIA, General Catalyst, Adobe Ventures, Figma Ventures, Canva and Deutsche Telekom’s T.Capital. Learn more at bfl.ai.

Media Contact
press@blackforestlabs.ai