AI Singapore & Alibaba unveil regional model for 1,200 languages

AI Singapore has launched the latest version of its large language model, Qwen-SEA-LION-v4, with technical support from Alibaba.

The model aims to address the diverse linguistic and cultural requirements of Southeast Asia, a region home to more than 1,200 languages. Built on Alibaba's Qwen3-32B foundation model, Qwen-SEA-LION-v4 is designed to improve multilingual accuracy and ensure regional accessibility in artificial intelligence applications.

Model performance

Qwen-SEA-LION-v4 currently leads the Southeast Asian Holistic Evaluation of Language Models (SEA-HELM) leaderboard among open-source models with fewer than 200 billion parameters. The model is noted for its multilingual support, advanced reasoning, and ability to understand long-context information specific to Southeast Asian languages. It has been optimised to operate on consumer-grade hardware, requiring only 32GB of RAM, which reflects a focus on accessibility for developers and enterprises.

Linguistic training

The Qwen3-32B base model underlying the new release was pre-trained on data spanning 119 languages and dialects, across 36 trillion tokens. This wide-ranging initial training includes significant coverage of Southeast Asian languages, which are often underrepresented in similar models. AI Singapore then further trained the model with over 100 billion tokens specific to Southeast Asian languages, enhancing its capacity to interpret local expressions, conversational subtleties and knowledge domains relevant to the region.

Commercial and social impact

There has been a noted lack of representation for Southeast Asian languages in global AI models, as most commercially available solutions prioritise English. The development of Qwen-SEA-LION-v4 seeks to address this imbalance by catering to the specific cultural and commercial needs of Southeast Asia's multilingual markets.

"Our collaboration with Alibaba on Qwen-SEA-LION-v4 is an important milestone in advancing AI inclusivity and to make it more representative of Southeast Asia. It embodies our shared vision of accelerating AI innovation across the region and ensuring that developers, enterprises, and public institutions have access to AI that is open, affordable, and locally relevant, and is designed to truly understand the languages, cultures, and communities of this region," said Dr Leslie Teo, Senior Director of AI Products, AI Singapore.

Collaboration details

Alibaba provided the Qwen3-32B foundation model and ongoing technical support for model training, while the team at AI Singapore contributed region-specific data for curation and evaluation. This joint approach enabled the resulting model to be both open-source and specifically tailored to Southeast Asian language tasks.

"Through this partnership with AI Singapore, we are proud to see the Qwen foundation model empowering the next wave of AI innovation in Southeast Asia. By combining our model's multilingual and reasoning strengths with AI Singapore's deep regional expertise, Qwen-SEA-LION-v4 demonstrates how open collaboration can make advanced AI more inclusive and locally relevant. We look forward to enabling more developers, enterprises and public-sector partners to build applications that truly understand the languages and cultures of this region," said Hon Keat Choong, General Manager of Singapore, Alibaba Cloud Intelligence.