MOSS-Image drew short onomatopoeia like "Dokan" (Boom) and speech bubble text specified in the prompt exactly as requested, even in Japanese. When I had Anima draw with the same prompt, it could draw ...