The Intelligent AI Assistant Doubao, developed by ByteDance, has recently upgraded its text-to-image capabilities. The new feature allows users to generate images containing specified text with a single click. The relevant person in charge of the Doubao large model team pointed out that the text-to-image model of Doubao integrates LLM (Large Language Model) and DIT (Digital Image Technology) framework, optimizing its learning ability for Chinese data, and enhancing the generation capability of Chinese characters, significantly improving the generation results.