According to China News, Ni Maizaxi, academician of the Chinese Academy of Engineering and professor at Tibet University, recently announced in an interview with China News that Tibet has made a major breakthrough in the research and development of Tibetan language large models, and has launched a trillion-parameter Tibetan foundation large model "Sunshine Qingyan" V1.0. This achievement marks that Tibet's artificial intelligence (AI) development has officially moved from simple scenario applications to a systematic research and development stage.
Currently, AI technology is widely applied in Tibet, covering areas such as government affairs, communities, public services, and ecological scientific research. For example, AI technology has been integrated into communities and airports in Lhasa, while the "e-Scientific Expedition" app has achieved rapid holographic interpretation and real-time data upload of research subjects during the Qinghai-Tibet scientific expedition. The State Council's "Opinions on Deeply Implementing the 'Artificial Intelligence+' Action" also provides a development blueprint for Tibet's AI.
The "Sunshine Qingyan" V1.0 model is a self-developed achievement of Tibet, trained on approximately 2.88 billion Tokens of high-quality Tibetan data, covering extensive fields such as law, medicine, and technology.
At the same time, Danzeng Luobo's team of young entrepreneurs in Tibet has been working on the digitalization of Tibetan language data since 2018, and has built nearly 70 million Tibetan-Chinese parallel corpora. The Tibetan large model developed by this team, DeepZang, has been integrated into the DeepSeek technology ecosystem, achieving high-quality multilingual semantic understanding and interaction, and is planned to be available to the public by the end of this year.
Tibet is actively seizing the AI high ground through the development of core base models, injecting strong momentum into regional economic and cultural development.
