DALL-E

OpenAI's text-to-image model, accessed primarily through ChatGPT and the OpenAI API

0Design & CreativePay Per Use

Editorial Review

DALL-E was the first AI image generator to capture mainstream attention (DALL-E 1 in early 2021, DALL-E 2 in 2022). In 2024 OpenAI consolidated image generation into ChatGPT and the API, retiring the standalone labs.openai.com product. The current model — accessed via ChatGPT or the gpt-image-1 endpoint — has caught up to and in some areas surpassed competitors on prompt adherence and in-image text rendering. For most users, the easiest access is through a ChatGPT Plus subscription ($20/mo), which includes generous image generation alongside text capabilities. Developers integrating into products use the API at roughly $0.04 per standard-resolution image. For free access, Microsoft's Copilot and Bing Image Creator wrap the same model with daily free generations. Compared to Midjourney (more aesthetic by default), Stable Diffusion (more controllable, local-first), and the latest Imagen/Veo from Google, DALL-E's particular strengths are prompt adherence, accurate text-in-image rendering, and seamless conversational editing inside ChatGPT — which makes it especially good for content workflows where the prompter is already in a chat session.

Key Features

High-quality text-to-image generationAvailable inside ChatGPT (Plus, Team, Enterprise)OpenAI API access (gpt-image-1 endpoint)

About DALL-E

DALL-E is OpenAI's text-to-image generation model. As of 2024, the standalone labs.openai.com product was sunset and image generation is now accessed through ChatGPT (Plus, Team, Enterprise) and the OpenAI API (gpt-image-1 endpoint). The current generation produces high-quality images from natural-language prompts and supports inpainting, image variations, and conversational image editing inside ChatGPT.