OpenAI has integrated a powerful new image generation tool into ChatGPT, leveraging its advanced GPT-4o model. This feature allows users to create images directly within the chatbot, marking a significant upgrade over previous reliance on DALL·E. As of March 25, 2025, the tool is available across multiple subscription tiers, including Plus, Pro, Team, and Free. The addition aims to enhance the AI-driven communication experience by offering more seamless and creative ways to generate visuals.
CEO Sam Altman celebrated the new feature, calling it a major technological advancement while acknowledging that it could result in content that may be controversial. He emphasized the importance of creative freedom and user control over generated content, expressing hope that the technology would be used responsibly. OpenAI plans to continue monitoring the societal impact of the tool as it evolves, particularly with respect to ethical concerns and AI’s role in creativity.
GPT-4o’s image generation capabilities surpass previous models with new features like improved text rendering, multi-turn generation, and better instruction following. Users can refine their images through ongoing conversations with the AI, allowing for more precision in tasks like character design and branding. Additionally, the model excels at incorporating detailed prompts, generating accurate depictions of complex scenes, and analyzing uploaded images to enhance new creations, making it a versatile tool for various creative processes.
Looking ahead, OpenAI will offer API access to developers, expanding GPT-4o’s image generation across different applications. The rollout of the feature also includes plans to make it available to Enterprise and Education users in the near future. Although rendering images may take up to a minute due to the model’s complexity, users can easily generate custom visuals by describing their needs in ChatGPT, with options for adjusting design elements like color and aspect ratio.
source: naira metrics