A Japanese-optimized large language model developed by NII. It is one of eight model sizes ranging from 150M to 172B parameters. Trained on the llm-jp-corpus v3 dataset, with SFT and DPO applied via Instruct2/Instruct3. Models under 13B are released under Apache 2.0 license.
Parameters
Undisclosed
Context Window
License
Proprietary
Release Date
2026-02-24
Japanese Language Capability
🇯🇵Native JP
Model developed by a Japanese company or specialized for Japanese. Highest Japanese understanding and generation capability.
API Pricing
API pricing for this model is not yet available
Strengths
- •Features an extended 32K token context window.
- •Combines MoE architecture with long-context processing.
- •Advanced instruction tuning for detailed task execution.
Weaknesses
- •Long context processing significantly increases memory and compute requirements.
- •Inference speed can be slower for very long input sequences.
- •Maintaining coherence over 32K tokens can be challenging.
Use Cases
- •Analyzing and summarizing lengthy Japanese documents.
- •Long-form dialogue systems and interactive storytelling.
- •Process documentation that requires understanding extended context.