gpt-image-1
2025-04-23 · Publisher : OpenAI · Domain : Artificial Intelligence
OpenAI makes gpt-image-1 available in the API, the native multimodal model that powers ChatGPT's image generation, to create and edit images with faithful text rendering and precise instructions.
What changes
- ChatGPT's image model becomes accessible to developers through the Images API, with support for the Responses API announced.
- Varied styles, adherence to custom instructions, world knowledge and accurate text rendering.
- Moderation parameter (auto by default or low) and C2PA metadata in every image.
- Already used by Adobe, Figma, Wix, HeyGen and others, according to the announcement; Canva and GoDaddy are experimenting with it.
- Organization verification may be required before access.
### Specifications - Text and image input, image output. - API identifier: gpt-image-1, image generation and image editing endpoints. - One month after its introduction in ChatGPT: more than 130 million users and more than 700 million images created in the first week. ### Pricing Announcement pricing per million tokens (USD): text input 5, image input 10, image output 40. Approximate cost per square image: about 0.02 (low quality), 0.07 (medium), 0.19 (high). The current documentation lists 0.011, 0.042 and 0.167 dollars for 1024x1024. ### Benchmarks No numerical score published by the vendor in the announcement. ### Key takeaways gpt-image-1 industrializes image generation in design, e-commerce and marketing tools. It was replaced at the top of the lineup by GPT Image 1.5 on December 16, 2025.
Links
Tags : Image, OpenAI, GPT-4o, API, Multimodal