Artificial Intelligence

Unveiling GPT-image-1: Rising to new heights with picture technology in Azure AI Foundry

April 24, 2025
roosho

We’re thrilled to announce the launch of GPT-image-1, the newest and most superior picture technology mannequin, now out there on Microsoft Azure OpenAI Service.

We’re thrilled to announce the launch of GPT-image-1, the newest and most superior picture technology mannequin, coming quickly to eligible prospects on Microsoft Azure OpenAI Service. This groundbreaking mannequin units a brand new customary in producing high-quality photographs, fixing advanced prompts, and providing zero-shot capabilities in varied situations.

That’s sufficient textual content to explain a picture technology mannequin for now—try the video under and see GPT-image-1 in motion:

Key options and enhancements

GPT-image-1 builds upon the strengths of its predecessor, DALL-E, with important enhancements:

  • Granular instruction response: GPT-image-1 excels at understanding and executing detailed directions, guaranteeing exact and correct picture technology.
  • Textual content rendering: The mannequin reliably renders textual content inside photographs, enhancing its utility in creating instructional supplies and storybooks.
  • Picture enter acceptance: Customers can add photographs and supply textual content prompts to generate new photographs or edit current ones, providing a flexible instrument for artistic tasks.
A virtual open house with a room with a couch and a window

GPT-image-1 capabilities

GPT-image-1 helps a number of modalities and options:

  • Textual content-to-image: Generate photographs from textual content prompts, just like text2im in ChatGPT DALL-E.
  • Picture-to-image: Create new photographs from user-uploaded photographs and textual content prompts, a characteristic not out there in ChatGPT DALL-E.
  • Textual content transformation: Edit photographs utilizing textual content prompts, akin to the rework characteristic in ChatGPT DALL-E.
  • Inpainting: Edit photographs with textual content prompts and user-drawn bounding bins, just like inpainting with DALL-E.
A group of images of people in a room

Use instances

GPT-image-1 is designed to energy a variety of functions, together with:

  • Academic materials technology: Create visible aids and interactive content material for studying.
  • Storybook creation: Generate constant and interesting illustrations for youngsters’s books.
  • Sport manufacturing: Develop sport property with constant type and character design.
  • UI designs: Design consumer interfaces with photorealistic components and coherent layouts.
A screenshot of a computer

Technical specs

  • Decision: Helps photographs with a minimal width and top of 1024 pixels, together with 1024×1024, 1024×1535, and 1535×1024 resolutions.
  • API integration: GPT-image-1 is accessible through API, within the Azure AI Foundry Picture Playground.

Security and moderation

GPT-image-1 is constructed with a sturdy security stack from OpenAI, together with c2pa and enter/output moderation. Azure AI particular gadgets embrace: Content material security and abuse monitoring.

Get began in the present day

Unleash your artistic potential with GPT-image-1, the cutting-edge expertise designed to raise your creative tasks. With capabilities that assist high-resolution photographs and seamless API integration, you may effortlessly deliver your visions to life. Expertise the beautiful photorealistic components and coherent layouts that can set your tasks aside. Use moral and secure picture technology with GPT-image-1’s sturdy moderation programs, making it the clear selection for all of your artistic wants.

Uncover the transformative energy of GPT-image-1 in the present day.