Black Forest Labs Launches FLUX 3 AI: Transitioning from Stills to Video with Robotic Integration

Black Forest Labs Launches FLUX 3 AI

Black Forest Labs (BFL) has announced the early access release of FLUX 3, a groundbreaking AI model capable of generating video clips up to 20 seconds long, complete with synchronized audio. This marks a significant shift for the company, which is primarily known for its image generation technology.

The innovative model uses a multimodal approach, allowing it to process images, video, and audio simultaneously. Initial reviews highlight that FLUX 3 has outperformed competitor models, with human evaluators favoring its output over Runway Gen-4.5 in 77% of tests and Luma Ray 3.2 in 93% of comparisons.

FLUX 3 not only excels in video generation but also continues to produce high-quality still images, showcasing its versatility across different artistic styles. Co-founder and CEO Robin Rombach emphasized the model's broader purpose, stating that understanding video helps it grasp essential physical principles necessary for robotic applications.

This is exemplified in the FLUX-mimic project, which is being tested by Audi. This initiative integrates FLUX 3’s capabilities with robotics to enhance automation in complex tasks, such as installing flexible door seals. Audi has reported that the new system can perform tasks that previous automation struggled with, responding in about 101 milliseconds, akin to human reflexes.

While FLUX 3 is currently accessible through APIs and selected partners, the open-weight “Dev” version is anticipated for release in late 2026. BFL's previous models have achieved notable success, outperforming other image generators and solidifying the company's standing in the AI landscape.