Create high-quality images in different styles with Kandinsky
The Kandinsky neural network is a project developed by Sber, which allows users to generate images based on text descriptions using neural networks. It is capable of generating images that are close to photorealism, and is based on Russian-language queries, which distinguishes it from most similar services.
Kandinsky uses a text query to generate images. For example, you can describe the image, set the style, and clarify details. The neural network has the ability to generate images of people, celebrities, various textures, as well as draw landscapes and animals in various styles.
In the latest version of the neural network, Kandinsky 2.1, 4 generation modes are available. In addition, the model can create images in various styles.
The service supports more than 100 languages, including Russian, and it is completely free.
You can use Kandinsky through the Fusionbrain website, Telegram bot or Rudalle website.
In October 2023, Kandinsky 2.2 added the ability to generate four-second videos using a text description.
Overall, Kandinsky is a powerful image generation tool that provides a lot of options for users.
The Kandinsky neural network was first published on 03-11-2023 20:56:21 and manually edited on 03-11-2023 21:15:59.