OpenAI just dropped a major update for ChatGPT users—real-time image generation powered by the GPT-4o model is now live. And yes, it’s available to everyone, including free-tier users. This marks a significant step forward in OpenAI’s efforts to bring multi-modal AI capabilities into everyday workflows, blending natural language, image creation, and code like never before.
What’s New?
The new image generation feature builds on the success of DALL·E 3 but brings faster, more accurate, and user-friendly performance. GPT-4o (short for “omnimodal”) can now turn detailed text prompts into high-quality images in under a minute—directly inside ChatGPT.
Users can control:
- Aspect ratio
- Color palettes
- Background transparency
- Scene composition
And it’s not just the Plus or Pro subscribers who get to play—Free, Team, and Pro users all have access (though free-tier users will encounter some usage limits).
Smarter Images, Fewer Mistakes
GPT-4o solves many of the classic AI art frustrations. It has stronger attribute binding (making sure things like “a red jacket on a tall man” don’t become “a red tall man”) and more accurate text rendering—so signs, logos, or in-image captions actually make sense.
Not Without Glitches
But of course, no launch is perfect.
Over the past 24 hours, users noticed that the AI had trouble generating certain types of requests. Specifically, it would create images for “sexy men” but refused prompts involving “sexy women.” This discrepancy sparked backlash online, with users questioning the model’s internal filters.
OpenAI CEO Sam Altman acknowledged the bug and confirmed that the issue was being investigated and would be fixed. He emphasized the balance between open creativity and responsible safeguards.
Ethics, Safety, and Ownership
OpenAI continues to focus on ethical deployment. Every generated image includes invisible digital watermarking to signal that it was created by AI. Despite this, users maintain full ownership of their generated images within OpenAI’s terms of service—great news for creators, marketers, and businesses using AI for branded visuals.
Watch It in Action
Curious about what it looks like? Here’s a recent demo of GPT-4o’s image generation in real-time, including how it integrates with Sora and other tools:
Final Thoughts
With this rollout, OpenAI is pushing the boundaries of what a single AI model can do. Whether you’re building a brand, visualizing an idea, or just having fun with prompts, GPT-4o’s real-time image generation feels like another step closer to creative AI becoming mainstream.
Stay tuned for more updates—and possible feature expansions—as GPT-4o continues to evolve.