Kandinskiy is Sber’s generative AI model for creating images and videos from text prompts. It understands prompts in Russian and English, supports different visual styles, and helps quickly draft visuals for design, content, presentations, and personal projects.
The current Kandinsky 5.0 line adds models for HD image generation and video generation from a text prompt or a starting frame. This makes Kandinskiy more than a text-to-image tool: it is now a broader visual AI platform inside the GigaChat ecosystem.
Kandinskiy is especially useful when you need Russian-language prompts, local cultural context, or Cyrillic text in generated visuals. For commercial use, check the usage rights and the terms of the platform where the model is available.
For adjacent image and video tools, compare Kandinskiy with Fusion brain and AI Video Generators.
