Singapore – AI Singapore is working with Dell Technologies to support the development and deployment of the SEA-LION family of open-source LLMs designed for SEA languages.
The collaboration involves testing and validating the models across a range of Dell AI PCs and edge systems, with the objective of enabling efficient performance in lightweight and resource-constrained environments.
LLMs developed at a global level have tended to prioritise English, limiting their effectiveness in capturing the linguistic diversity and cultural context across the region. SEA-LION seeks to address this limitation through training on a substantial corpus of region-specific data, allowing the models to reflect local language patterns and usage.
As an open-source initiative, the project lowers barriers associated with the high costs of developing foundational models, supports regional ownership of AI capabilities and offers an alternative to proprietary platforms through permissive licensing that encourages community participation.
“This collaboration is a key step in realising our vision for SEA-LION–models that are not only local and culturally relevant but also resource-efficient enough to be deployed on the edge,” Dr Leslie Teo, senior director of AI Products at AI Singapore, stated.
“By running fully featured LLMs directly on Dell devices, we provide enterprises with privacy and lower cost alternatives while remaining responsive and reliable.”
Dell Technologies contributes infrastructure and optimisation expertise to support the deployment of quantised models such as SEA-LION on edge devices. Its AI-enabled personal computers and edge platforms are designed to deliver consistent performance while minimising resource requirements, supporting use cases in environments with limited connectivity or computing capacity.
One practical application of the collaboration is AI Singapore’s voice transcription system for SEA languages, which integrates automatic speech recognition technology with the SEA-LION model and operates on Dell AI PCs. The system enables real-time transcription for applications such as chatbots, reporting tools and multilingual communication services.
By allowing large language models to run directly on local devices, the solution supports offline operation, data privacy and low-latency performance for enterprise use.
“SEA-LION’s transformative understanding of 11 Southeast Asian languages and cultural nuances requires practical application,” Andy Sim, vice president and managing director of Singapore at Dell Technologies, shared.
“This collaboration demonstrates that this sophisticated, culturally intelligent AI can run efficiently on laptops and edge devices. Together, we are democratising AI in Southeast Asia and are fostering a future where advanced AI is accessible to all.”
Looking ahead, AI Singapore plans to further develop the SEA-LION models by extending their capabilities in audio and speech processing, improving linguistic precision and enhancing efficiency to support deployment across a broader range of hardware platforms.
The initiative is also intended to provide a foundation for future agent-based AI systems capable of handling complex tasks across multiple sectors and languages within SEA.
