DALL·E 3

OpenAI's advanced text-to-image model, DALL-E 3, generates high-fidelity visuals with superior prompt adherence and the unique capability to render legible text within the image.

DALL-E 3 is OpenAI’s third-generation text-to-image AI system, launched in October 2023 for ChatGPT Plus and Enterprise users. This model marks a significant advancement over DALL-E 2, largely due to its deep integration with GPT-4: the language model automatically refines and expands vague user requests into detailed prompts, ensuring greater accuracy and nuance in the final output. It excels at following complex instructions, producing images with exceptional detail, striking compositions, and crucial support for multiple aspect ratios (e.g., 1024x1024, 1024x1792). Key capabilities include generating sharp, readable text within the image, a major differentiator in the AI art space.

https://openai.com/index/dall-e-3/