The Ukrainian large language model will be built on Llama 4, as announced by Deputy Minister of Digital Transformation Oleksandr Bornyakov during the 2U Tech Forum. According to him, the first version may appear by the end of the year and will be based on 2–3 billion tokens.
The next step will be expanding the model to 11 billion tokens.
"In general, we want to scale the model to 70 billion tokens. We want it to be used not only for our internal tasks but also for any company in Ukraine to utilize it," added Bornyakov. He mentioned that the roadmap for the model's launch is already planned, and in about a month, announcements will be made about where and how it will take place.
In early February, the Ministry of Digital Transformation launched the AI Center of Excellence, the first of its kind in Ukraine, which is also working on the development of the Ukrainian large language model (LLM).
The development of the national LLM will pave the way for the mass creation of Ukrainian AI tools. Based on this model, thousands of products will emerge, such as virtual assistants like ChatGPT, text generators like Notion AI, and programming tools similar to GitHub Copilot, as shared by him at the Forbes AI Day 2025.
Fedorov emphasized that having a domestic LLM model will enhance national security, as foreign models may be influenced by Russian propaganda and may not align with the Ukrainian context. Local data processing will become strategically important for AI in sectors like defense, government, healthcare, and finance.
He added that countries are increasingly developing their own LLMs in national languages, so Ukraine must act quickly to not only catch up with the leaders but also become an independent player in technological competition. The launch of the model is a key task for the WINWIN AI Center of Excellence, scheduled for the end of 2025.
