Launched in March 2025 in ChatGPT, the image generation model gpt-image-1 has seen a huge surge in popularity: over 700 million images were created in one week, which led OpenAI to limit access to this feature. This native multimodal model, capable of producing images in a wide variety of styles, following detailed instructions, integrating text, and contextualizing content, is now available to developers and businesses through the Image API.
According to OpenAI, the versatility of the model unlocks countless practical applications in various fields. Several companies have already integrated the model to enhance their offerings:
 
  • Creative Tools: Adobe, Figma, and Canva are leveraging GPT-Image-1 to enrich their design platforms and offer advanced image generation and editing features;
  • E-commerce: GoDaddy integrates the model to help users create customizable logos and marketing content tailored to their visual identity;
  • Marketing and Communication: HubSpot is exploring the use of GPT-Image-1 to generate professional visuals for social media, ad campaigns, and landing pages;
  • Education: Educational platforms are using the model to create interactive and illustrated teaching materials;
  • Video Games: Some companies in the gaming industry are using GPT-Image-1 to generate graphic assets.

Pricing

Access to gpt-image-1 via the API is based on differentiated token pricing: $5 per million text tokens, $10 for image inputs, and $40 for generated images. In practice, this translates to a cost between $0.02 and $0.19 per generated image depending on quality. This pricing structure encourages gradual adoption, especially for applications requiring high usage frequency and advanced customization.
On the security side, OpenAI applies the same safeguards as for image generation in ChatGPT: integration of C2PA metadata to identify images generated by the model, configurable moderation, and no use of client data for training.
Currently available through OpenAI's Image API, GPT-Image-1 will soon be integrated into the API Responses, thereby expanding possible interaction modalities for developers.

Translated from OpenAI intègre GPT-Image-1 à son API Images

To better understand

What is the C2PA standard and how does it apply to GPT-Image-1?

The C2PA (Coalition for Content Provenance and Authenticity) standard is a protocol to authenticate the origin of digital content, like AI-generated images. In GPT-Image-1, it is used to embed metadata that identifies the created images, ensuring transparency and combating misinformation.