Event ·Multi-source

FLUX 3 发布：Black Forest Labs 推出真实世界模型

Black Forest Labs 发布 FLUX 3，定位为真实世界模型，统一处理图像、视频、音频三种模态。模型最长生成约20秒带原生音频的视频，支持续写、关键帧、多语言对白。图像合成与编辑功能将在数周后推出，强化复杂 prompt 和文字渲染。动作预测已通过 FLUX-mimic 用于 mimic robotics 和 Audi 场景。

Current read

FLUX 3 把图像、视频、音频放一起训，还能预测动作。已经能生成20秒带声音的视频，图像也快上了。

9 sources72° AI heatLast updated 2026/7/24 03:44:57

Evidence

9 sources

Primary sourceshao__meng

2026/7/24 03:44:57View source

Related source 1berryxia

2026/7/23 23:23:45View source

Related source 2Latent Space (swyx)

2026/7/24 04:30:12View source

Related source 3歸藏(guizang.ai)

2026/7/23 15:43:23View source

Related source 4Justine Moore

2026/7/23 15:47:25View source

Related source 5Decoder

2026/7/23 18:03:01View source

Related source 6IT之家

2026/7/24 06:48:04View source

Related source 7@koltregaskes

2026/7/23 22:58:45View source

Related source 8岚叔

2026/7/24 00:20:50View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story