Thailand Launches Initiative for Local Large Language Model Development
The National Science and Technology Development Agency partners with Siam AI Corporation to enhance the country’s artificial intelligence capabilities.
The National Science and Technology Development Agency (NSTDA) of Thailand has entered into a partnership with Siam AI Corporation to develop a Thai Large Language Model (LLM).
This initiative represents a significant step in enhancing the nation's artificial intelligence (AI) capabilities and aligns with its sovereign AI strategy.
The partnership is supported by an investment of 300 million baht from Siam AI.
During the memorandum of understanding (MoU) signing ceremony, Sukit Limpijumnong, president of NSTDA, emphasized the importance of a Thai LLM that understands local context and culture.
He described the project as a crucial component for Thailand’s digital infrastructure, aimed at fostering growth in AI development specifically tailored to the Thai language.
The implementation of the Thai LLM is expected to play a vital role in advancing AI technologies to meet national needs, driving innovation across digital services, and enhancing the competitiveness of the country.
NSTDA plans to conduct deep research and development initiatives, including the establishment of a sandbox for proof of concept, and to cultivate a skilled workforce in collaboration with Siam AI to strengthen the research ecosystem in the long term.
Ratanaphon Wongnapachant, chief executive of Siam AI, stated that the collaboration serves as a foundational step for developing a Thai LLM for commercial applications in both public and private sectors.
He highlighted the necessity of developing native Thai models, claiming that reliance on foreign-developed models could lead to the loss of valuable data integral for cognitive development, which is essential for the country’s growth.
Large language models with multimodal capabilities are significant for progressing next-generation AI applications, such as autonomous vehicles, which require the integration of various data types including text, vision, audio, and sensor information for real-time decision-making.
Siam AI boasts one of the most advanced high-performance computing infrastructures in Asia, featuring Nvidia’s Blackwell architecture and liquid-cooled rack clusters specifically designed to support LLMs.
Thailand is among only four Asian countries—along with China, Japan, and South Korea—that have developed their own LLMs. The Thai LLM is projected to encompass 7 billion parameters and is expected to be completed within three months.
The current initiative builds upon a prior funding phase that received 20-40 million baht.
Chai Wutiwiwatchai, executive director of the National Electronics and Computer Technology Center (Nectec), noted that Nectec’s expertise in multimodal LLMs will be instrumental in supporting Siam AI, which has shown considerable capability in text-based LLM development.
Nectec is responsible for developing Pathumma LLM, tailored for multimodal data processing to handle state and public sector documents while focusing on official language and its contextual nuances.
Additionally, Nectec plans to deploy government AI projects utilizing Siam AI's infrastructure, which will facilitate scalability and flexibility for pilot initiatives.
The agency also aims to develop a skilled workforce in high-performance computing critical for future advancements.
According to the Oxford Insights Government AI Readiness Index 2024, Thailand ranks 35 out of 188 countries in terms of government readiness for AI.
Sukit noted that Thailand is demonstrating strong momentum in AI development, particularly in the medical sector, where the country has established expertise.
He highlighted the potential for Thailand to export AI-driven services in the future.
In May, Siam AI formalized an agreement with Naver Cloud, a South Korean firm specializing in cloud computing and AI solutions, to co-develop a Thai LLM and AI agents aimed at the tourism sector.
According to Naver Cloud, this collaboration will leverage extensive Thai language datasets and expertise in LLM development to produce a Thai-specific model by the end of the year, designed for practical applications within the high-demand tourism industry.
This initiative represents a significant step in enhancing the nation's artificial intelligence (AI) capabilities and aligns with its sovereign AI strategy.
The partnership is supported by an investment of 300 million baht from Siam AI.
During the memorandum of understanding (MoU) signing ceremony, Sukit Limpijumnong, president of NSTDA, emphasized the importance of a Thai LLM that understands local context and culture.
He described the project as a crucial component for Thailand’s digital infrastructure, aimed at fostering growth in AI development specifically tailored to the Thai language.
The implementation of the Thai LLM is expected to play a vital role in advancing AI technologies to meet national needs, driving innovation across digital services, and enhancing the competitiveness of the country.
NSTDA plans to conduct deep research and development initiatives, including the establishment of a sandbox for proof of concept, and to cultivate a skilled workforce in collaboration with Siam AI to strengthen the research ecosystem in the long term.
Ratanaphon Wongnapachant, chief executive of Siam AI, stated that the collaboration serves as a foundational step for developing a Thai LLM for commercial applications in both public and private sectors.
He highlighted the necessity of developing native Thai models, claiming that reliance on foreign-developed models could lead to the loss of valuable data integral for cognitive development, which is essential for the country’s growth.
Large language models with multimodal capabilities are significant for progressing next-generation AI applications, such as autonomous vehicles, which require the integration of various data types including text, vision, audio, and sensor information for real-time decision-making.
Siam AI boasts one of the most advanced high-performance computing infrastructures in Asia, featuring Nvidia’s Blackwell architecture and liquid-cooled rack clusters specifically designed to support LLMs.
Thailand is among only four Asian countries—along with China, Japan, and South Korea—that have developed their own LLMs. The Thai LLM is projected to encompass 7 billion parameters and is expected to be completed within three months.
The current initiative builds upon a prior funding phase that received 20-40 million baht.
Chai Wutiwiwatchai, executive director of the National Electronics and Computer Technology Center (Nectec), noted that Nectec’s expertise in multimodal LLMs will be instrumental in supporting Siam AI, which has shown considerable capability in text-based LLM development.
Nectec is responsible for developing Pathumma LLM, tailored for multimodal data processing to handle state and public sector documents while focusing on official language and its contextual nuances.
Additionally, Nectec plans to deploy government AI projects utilizing Siam AI's infrastructure, which will facilitate scalability and flexibility for pilot initiatives.
The agency also aims to develop a skilled workforce in high-performance computing critical for future advancements.
According to the Oxford Insights Government AI Readiness Index 2024, Thailand ranks 35 out of 188 countries in terms of government readiness for AI.
Sukit noted that Thailand is demonstrating strong momentum in AI development, particularly in the medical sector, where the country has established expertise.
He highlighted the potential for Thailand to export AI-driven services in the future.
In May, Siam AI formalized an agreement with Naver Cloud, a South Korean firm specializing in cloud computing and AI solutions, to co-develop a Thai LLM and AI agents aimed at the tourism sector.
According to Naver Cloud, this collaboration will leverage extensive Thai language datasets and expertise in LLM development to produce a Thai-specific model by the end of the year, designed for practical applications within the high-demand tourism industry.
