gpt-image-1.txt•813 B
OpenAI just introduced gpt-image-1, the latest AI model designed to generate images from plain text prompts. This update is now available through both OpenAI’s own API and Azure AI Foundry, making it even easier for anyone to experiment with image generation. To test its abilities, the team generated pixel art cats with the same prompt across three models—DALL·E 2, DALL·E 3, and gpt-image-1—highlighting how each interprets the same creative task. The results show different visual styles and a unique flair from the new gpt-image-1, which brings fresh options for creators seeking distinctive, high-detail pixel art. With this launch, you can now compare outputs side by side, observing nuanced changes in how these models translate words into images—and discover which one fits your next idea best.