Kazakh Tech Firm Introduces First Large Language Model for Eastern Armenian

ASTANA – Kazakh technology company Gen2B has launched HyGPT, the first large language model (LLM) designed for the Eastern Armenian language, reported the company’s press service on July 1.

Photo credit: istockphoto.com

Developed in collaboration with the National Center for Artificial Intelligence and Technologies of Armenia (NCCAIT), the model is based on the Gemma 2 architecture and trained on 10 billion Armenian tokens. It demonstrates strong performance in translation, question answering, mathematical reasoning, and instruction following.

“We now clearly understand how to monetize vertical Artificial Intelligence (AI) bots for large and mid-sized companies, especially in countries with low-resource languages. We are ready to help businesses boost sales through AI as part of a new and emerging economy,” said Bakht Niyazov, the co-founder of Gen2B.

However, Niyazov emphasized that the company’s core philosophy is to first create something useful before pursuing profit. 

“That’s why our first step in Kazakhstan was developing the Irbis language model. Now our focus is on Armenia, where we’ve built a similar system,” he said.

Gen2B plans to expand to Uzbekistan next, calling it a “promising market” for language model applications.

HyGPT is available for free download on the Hugging Face platform. It can be used in education, digital services, public applications and business.

Gen2B’s solutions are powered by an AI-driven platform that functions autonomously, analyzing, learning, and providing strategic recommendations based on data from communication channels. While Kazakh and Armenian models are free for both personal and commercial use, banks, telecom companies, and financial institutions must obtain permission from the company to access them.


Get The Astana Times stories sent directly to you! Sign up via the website or subscribe to our X, Facebook, Instagram, Telegram, YouTube and Tiktok!