精选理由
FLUX 3把图、视频、音频、机器人全做到一个模型里了，还已经在奥迪工厂跑起来了，视频版还能抢先体验。
FLUX 3是一个单一模型，同时支持图像、视频、音频生成以及机器人动作预测。其衍生版本FLUX-mimic已在奥迪工厂用于布线、分拣和装配。FLUX 3 Video现已开放早期访问，可加入候补名单。未来数周至数月内将推出原生音频、动作预测、图像生成、快速变体及开放权重骨干。
原文 · @koltregaskes
FLUX 3 is one model trained for image, video, audi…
FLUX 3 is one model trained for image, video, audio and action-prediction for robotics.
FLUX-mimic is already running on factory robots at Audi doing wiring, sorting and assembly work.
FLUX 3 Video early access is live now, check the waitlist. Native audio, action prediction, image generation, fast variants and open-weight backbone coming over the next weeks and months.
https://t.co/0lTuCsuzrK