Comprehensive Vision Language Model