The Institute of Smart Systems and Artificial Intelligence (ISSAI) has made a remarkable advancement in the field of AI with the launch of Oylan2. Oylan2 is a state-of-the-art language-audio-vision model capable of interacting in Kazakh, Russian, and English. 

ISSAI has undertaken extensive work to enhance the functionality of Oylan, introducing significant improvements to the performance and capabilities of the existing model. These advancements mark an important step forward in refining multimodal AI systems tailored to the linguistic and cultural context of Kazakhstan. The launch of Oylan2 is not just a technical upgrade—it is a testament to ISSAI’s unwavering dedication to innovation and excellence in the field of artificial intelligence.

This milestone represents yet another pivotal chapter in ISSAI’s ambitious journey to advance AI in Kazakhstan and the broader region. With Oylan2, Kazakhstan joins a select group of countries with this advanced technical expertise.

Compared to our previous models, KAZLLM and Oylan, Oylan2 delivers significantly better text processing and produces far fewer hallucinations. It also excels in image-related tasks such as image captioning, optical character recognition, visual-question answering, and chart-table understanding with much higher accuracy than its predecessor.

As one of the first publicly available language-audio-vision models in the world, Oylan2 also boasts exceptional audio processing abilities in Kazakh, Russian, and English. It can do automatic speech recognition, speech-to-text translation, language identification, music captioning, sound classification and much more.

As long as ISSAI can secure resources, the Institute will remain a strong competitor in the global AI race and will create advanced AI models, tools and services for the people of Kazakhstan.

For a detailed demo of Oylan2’s capabilities, please watch our extended YouTube video.

You can also test Oylan2’s advanced features at ISSAI Playground  (oylan.nu.edu.kz).