A compact Japanese-language specialist built on Qwen's 7B foundation, this model carries ELYZA's tuning work aimed at improving Japanese text understanding and generation. It handles everyday Japanese tasks with reasonable fluency for its size, though the 7B parameter count means it trades raw capability for accessibility and speed. It fits comfortably in resource-constrained environments where a full-scale model would be impractical.