OpenAI has unveiled a new image-generation feature within ChatGPT, powered by its latest model, GPT-4o.
The rollout, which began on March 25, is available across Plus, Pro, Team, and Free subscription tiers.
Previously, users had to rely on DALL·E for image creation, either as a standalone tool or within ChatGPT. Now, OpenAI has fully integrated advanced image-generation capabilities into ChatGPT, making it more seamless and interactive.
Sam Altman, OpenAI’s CEO, shared his thoughts on the development via his X page, describing it as “an incredible technology/product” that pushes the boundaries of creative freedom. He acknowledged that while the tool will empower users to create remarkable content, some outputs may be controversial.
“Two things to say about it: 1. It’s an incredible technology/product. I remember seeing some of the first images come out of this model and having a hard time, they were really made by AI.
“We think people will love it, and we are excited to see the resulting creativity. Secondly, this represents a new high-water mark for us in allowing creative freedom.
“People are going to create some really amazing stuff and some stuff that may offend people; what we’d like to aim for is that the tool doesn’t create offensive stuff unless you want it to, in which case within reason it does.
“As we talk about in our model spec, we think putting this intellectual freedom and control in the hands of users is the right thing to do, but we will observe how it goes and listen to society.
“We think respecting the very wide bounds society will eventually choose to set for AI is the right thing to do, and increasingly important as we get closer to AGI. Thanks in advance for the understanding as we work through this,” Altman stated.
Unlike earlier models, GPT-4o’s image-generation capabilities have been enhanced for greater precision and flexibility.
OpenAI highlighted several key improvements:
– Text rendering: The model can now generate clear and readable text within images, making it more effective for creating infographics, diagrams, and labeled visuals.
– Multi-turn generation: Users can refine images through conversational adjustments, ensuring consistency across different versions—ideal for character design, branding, and storyboarding.
– Instruction following: GPT-4o can process highly detailed prompts, accurately rendering images with multiple objects and maintaining spatial relationships.
– In-context learning: The AI can analyze and incorporate elements from uploaded images into new creations, making it useful for design inspiration and brainstorming.
– Knowledge integration: The model links text and visual understanding, enabling it to create context-aware visuals such as weather infographics, educational illustrations, and technical diagrams.
In addition to the ChatGPT rollout, OpenAI announced that developers will soon gain API access, allowing them to integrate GPT-4o’s image-generation capabilities into their own applications.
ChatGPT users can now generate images simply by describing their requirements, specifying colors, aspect ratios, and other design elements. However, due to the complexity of the model, OpenAI noted that rendering an image may take up to a minute.
For businesses and developers seeking more customization, OpenAI confirmed that DALL·E will still be available as a standalone model option.
With this latest development, OpenAI continues to push the boundaries of AI-driven creativity, ensuring users have greater control and flexibility in their image-generation experiences.
