OpenAI now offers ChatGPT’s Image Generation as an API

Now people can natively incorporate Studio Ghibli inspired pictures generated by ChatGPT inthe businesses they run. Openai added the model that powers its popular image-generation tool, ChatGPT, into its API.

Developers and enterprises can integrate high-quality professional-grade images directly into their tools and platforms using the gpt image-1 model. blog post.

Pricing of the API separates tokens by text and image. Text input tokens (or the prompt text) will cost $5 for 1 million tokens. Image input tokens are $10 per million tokens. Image output tokens or the generated image will cost $40 per million tokens.

Competition like Stability AI offers a Credit-based system for API with one credit equal to $0.01. Using its flagship Stable Image Ultra costs eight credits per generation. Google’s image generation model, Imagen, Charges paying users $0.03 for each image generated by the Gemini API.

Image generation in a single place

OpenAI enabled ChatGPT users in April to create and edit images directly in the chat interface, just a few months after integrating image generation into ChatGPT via the GPT-4o Model.

According to the company, image generation on the chat platform has “quickly become one of our most popular” features. OpenAI reported that over 130 million users accessed the feature in the first week and created 700,000,000 photos. OpenAI faced some challenges as a result of this popularity

. Social media users discovered that they could use ChatGPT to create images inspired by Japanese animation giant Studio Ghibli. As a result, all of my social media feeds for the weekend were filled with the exact same photos. OpenAI CEO Sam Altman was prompted by the trend. OpenAI added its image model DALL E 3 to ChatGPT previously. This model was a diffusion transform model, not the multimodal understanding native to GPT-4o.

Enterprise use cases

Many enterprises want to be able to generate images for projects and don’t want a separate app to do so. OpenAI’s API now includes the image model, allowing enterprises to integrate gpt-1 with their own ecosystems.

OpenAI stated that it has already seen several startups and enterprises use the model to create creative products, experiences and projects. It named several well-known companies in its blog.

Canva reportedly is exploring ways to integrate gpt image-1 for its Canva and Magic Studio Tools. GoDaddy has begun experimenting with image creation for customers to create logos. Airtable allows enterprise marketing and creative team to manage asset workflows easily at scale.

OpenAI stated that gpt image-1 will receive the same safety guardrails in the API as ChatGPT. The company said that images generated by the model include metadata from the Coalition for Content Provenance and Authenticity, (C2PA), which labels content as AI-generated and tracks owner. OpenAI is a member of C2PA’s steering committee. Users can also control the content moderation in order to generate images that are most aligned with their brand. OpenAI has promised not to use any customer API data including images generated or uploaded by gpt image-1 in order to train their models.