For the complete documentation index, see llms.txt. This page is also available as Markdown.

SEA-LION v4.8 (Latest)

SEA-LION version 4.8, released in Sep 2026, is our latest collection of foundational, text generation optimized for Southeast Asia. Utilizing open dataset for continue pretraining and advanced post-training methodologies—such as agentic on-policy distillation delivers state-of-the-art regional fluency and high computational efficiency.

Nemotron-SEA-LION-v4.8

Built on the highly efficient NVIDIA Nemotron-3 Nano (30B-A3B) and Super (120B-A12B) base architectures, our models are optimized for precise text generation while maintaining low active-weight overhead.

  • Nemotron-SEA-LION-v4.8

    Refer to the detailed page to access model cards, usage scripts, and deployment configurations for the entire nemotron family:

    • Focused on parallel-only dataset for the core text-generation Foundations

Last updated