The problem of creating a generative factographic information system with generative block has been investigated in the paper. The main aim of this paper is to present a new structural diagram of the generative factographic information system that could help scientists and researchers. Patents for inventions (device) and invention patents (software) of Russian Federation were analyzed and have found inventions applying modern generative technologies. Important ideas for modern generative tools and methods from these patents can be applied for generative block. Contemporary programming languages for instance Kotlin, C#, Java and Python, are uses in this software. Necessary set of blocks for the generative factographic information system with generative block for recognizing specified objects was described. Factographic information system composed of five blocks, four factographic database, special library of generative tools and methods and graph tools and methods. The paper adds to the consideration of the structure of the factographic information system. Python programming language with library PyTorch was selected because of its advantages over other programming languages. Key element of the generative factographic information system with generative block was successfully developed on Python language. Finally, the main results are outlined.