🎬 Kandinsky 6.0 Video: Create Videos with Sound

Kandinsky 6.0 Video — a new open-source lineup (MIT license): the flagship Pro and the lightweight Lite. Both generate videos with synchronized speech, music, and ambient sounds, in quality up to Full HD.

✅ How to use:
1. Describe a scene in text
2. OR upload first frame + audio description
3. Get 5-second video with sound

🎯 Prompt examples:
• "A cat walking on snow with crunching sounds"
• "Spanish guitar playing in a sunset scene"
• "Person saying 'Hello, welcome to our store'"
• "ASMR video with paper rustling and firewood crackling"
• "Chase scene with dramatic soundtrack"

🔥 Features:
• Lip-sync: lips move with speech
• Realistic physics
• 44 kHz audio quality
• Resolutions: SD, HD, Full HD
• Free demo on Hugging Face Spaces

📊 Performance:
• Pro clearly outperforms the previous version
• Pro beats Veo 3.1 Fast in image animation

🛠 For developers:
Open-source (MIT license). Available on Diffusers, FastVideo, ComfyUI, SGLang, vLLM-omni.

🔗 Hugging Face