Which OpenAI text-to-image model was released in April 2022?

OpenAI released DALL-E 2, its second major text-to-image model, in April 2022.

DALL-E 2 generated images from natural-language descriptions and could also create variations or edit portions of existing images. It followed the original DALL-E, which OpenAI announced in January 2021, and produced higher-resolution, more photorealistic results than its predecessor.

The model used a combination of diffusion-based image generation and learned text-image representations. Its public rollout began with a beta access program before broader availability through OpenAI’s services.

DALL-E 2 is often confused with DALL-E 3, which OpenAI introduced later in 2023. It is also distinct from GPT-4, a multimodal language model, and Sora, OpenAI’s later text-to-video model.