Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation

Kandinsky 6.0 Video introduces a dual-stream CrossDiT architecture for synchronized text-to-audio-video generation. The family includes a 3B parameter Lite model and a 29B parameter Pro model, both producing 5-second clips with 44 kHz audio and lip-syncing via bidirectional cross-attention. Integrated super-resolution scales outputs to 1920x1080 resolution.