Black Forest Labs Launches FLUX 3 Multimodal AI With Video and Audio Generation, But Limited Access Clouds Enterprise Appeal

Black Forest Labs launches FLUX 3, a groundbreaking multimodal AI capable of generating images, video with audio, and robotic action predictions from one unified model, but its gated early access, absent pricing, and lack of production SLAs leave enterprise buyers unable to fully evaluate it against already-available rivals like Google's Gemini Omni Flash.

• Black Forest Labs launches FLUX 3, a multimodal AI model capable of generating images, 20-second video clips with native audio, and robotic action predictions from a single unified architecture, marking the company's first public video generation model.
• FLUX 3 enters a limited gated early access program with no public API availability, no published pricing, no production SLAs, and only preliminary benchmark results, making it difficult for enterprise buyers to evaluate total cost of ownership or independently verify performance claims.
• In head-to-head preference testing, FLUX 3 outperforms several competitors including Luma Ray 3.2 and Runway Gen-4.5, but ties Google's Gemini Omni Flash at 52%, which is already generally available at $0.10 per second of generated video, putting FLUX 3 at a competitive disadvantage until pricing and full availability are announced.