Sberbank’s Kandinsky 2.1 neural network demonstrated what a Russian dream cottage could look like, as reported by the bank’s press service.
The images emerged from the most popular requests on Domclick, the leading real estate service in Russia, known for showcasing homes online.
The artificial intelligence analyzed the top requests from Domclick users and produced three visuals: a seaside house, a cottage village along the Moscow Ring Road, and a forested village.
Domclick analysts observed that, ahead of the summer season, interest in suburban real estate rises. They studied where residents of the Moscow region and Krasnodar Territory showed the strongest appetite for cottages in March 2023, outlining the patterns that shape demand across different districts.
The most sought after regions in the Moscow area include Ramenskoye, Domodedovo, Dmitrov, Naro-Fominsk, Chekhov, and Solnechnogorsk. Typical prices for a standard house, roughly 180 square meters, hover in the 11 to 15 million rubles range.
In Krasnodar Krai, buyers also weigh options in suburban towns with average prices between 9 and 14 million rubles. The cities drawing the most attention are Novorossiysk, Armavir, Yeysk, Anapa, and Temryuk.
Previously, Sberbank introduced Kandinsky 2.1, a neural network capable of turning natural language prompts into high quality images within seconds. It can blend multiple drawings, revise them via text instructions, generate visuals similar to a given image, fill in missing portions of a picture, and produce artwork within a limitless canvas mode that supports both inside and outside painting styles.