Qwen-SEA-LION-v4-32B-IT: Southeast Asian Language LLM

Developed by AI Singapore, Qwen-SEA-LION-v4-32B-IT is a 32 billion parameter instruction-tuned model built upon the Qwen3 architecture, featuring a 32,768 token context length. This model is part of the SEA-LION (Southeast Asian Languages In One Network) collection, specifically designed to enhance language understanding and generation for the Southeast Asian region.

Key Capabilities & Features

Multilingual Proficiency: Continued pre-training on approximately 100 billion tokens from the SEA-Pile v2 corpus, covering 7 key Southeast Asian languages: Burmese, Indonesian, Malay, Filipino, Tamil, Thai, and Vietnamese, in addition to English.
Instruction Following: Post-trained on 8 million high-quality question-and-answer pairs for robust instruction-tuning and multi-turn chat capabilities.
Qwen3 Foundation: Inherits strong foundational capabilities and support for over 100 languages from its Qwen3-32B base model.
Evaluated on SEA-HELM: Performance assessed across various tasks including QA, Sentiment Analysis, Translation, and more, using localized benchmarks like SEA-IFEval and SEA-MTBench.
"Thinking Mode" Feature: Includes an enable_thinking parameter for potentially enhanced reasoning processes during generation.

Ideal Use Cases

Applications requiring strong performance in Southeast Asian languages.
Instruction-following and multi-turn conversational agents for regional contexts.
Research and development in multilingual NLP, particularly for SEA languages.

Note: The model has not been aligned for safety, and users should implement their own safety measures. It is text-only, with vision capabilities comparable to Qwen3-32B.