Kandinsky 6.0 Video launches: open foundation models for synced video and audio generation

q5sys · reddit · 2026-10-08

Kandinsky Lab has released Kandinsky 6.0 Video, open foundation models for synchronized video and audio generation, with weights available on Hugging Face in diffusers format. They also shipped a companion video upscaling (VSR) model. The poster found v5 underwhelming and hopes v6 improves.

Related event: Kandinsky Lab Open-Sources Kandinsky 6.0 Video with Joint Audio-Video Generation under MIT License(7 posts)→

Original post →

More from Multimodal

• Running TRELLIS.2 Image-to-3D Reliably on a 10GB GPU: Open-Source CLI and the Settings That Fit — hamper10 · 2026-10-08
• Video Editing Is Going Through What Coding Did, Says Early Tester — ayushtweetshere · 2026-10-08
• Same Character, Every Shot: AI Video Demo Showcases Character Consistency — no3us · 2026-10-08
• India's IICT and YouTube premiere first four shorts from Create with AI programme — lmoroney · 2026-10-08
• Lora Pilot demo keeps the same character consistent across every video scene — no3us · 2026-10-08
• Vidu Q4 Preview debuts at No.3 on Artificial Analysis image-to-video leaderboard — ArtificialAnlys · 2026-10-08

Multimodal channel →