
ChatGPT has just received a major upgrade.
While it was already great at generating text, summarizing content, or chatting with you, image creation and editing required a separate model: DALL·E.
But now, OpenAI’s core model, GPT-4o, can natively generate and edit images and photos. That means ChatGPT can now handle everything — speaking, writing, drawing, and editing — all within one intelligent system.
This marks a major leap in how we interact with AI.
What’s Different This Time?
Previously, when you asked ChatGPT to generate an image, it relied on an external model, DALL·E 3. But now, GPT-4o itself can directly handle image creation and editing — no handoffs required.
Here’s a quick comparison:
| 🧩 Feature | Before (DALL·E 3) | Now (GPT-4o) |
| 🎨 Image Creation | External model | Handled directly by GPT-4o |
| ✂️ Editing Tools | Limited | Includes editing of backgrounds and people |
| 🔍 Accuracy | Good | Even more precise and realistic |
| ⏱️ Speed | Faster | Slightly slower but higher quality |
One highlight: GPT-4o can now edit images of people, or even fill in missing background and foreground elements — a big step forward.
Who Can Use It?
The feature is already live for users on the $200/month Pro plan. OpenAI has announced it will roll out to Plus users, free users, and API developers in the near future.
Also, this capability is now included in OpenAI’s AI video generation tool, Sora.
Respecting Artists’ Rights
OpenAI says GPT-4o was trained using a mix of publicly available data and licensed data through partners like Shutterstock. The company also enforces policies that prohibit generating images that imitate the work of living artists.
Additionally, creators can request to opt out of having their content used in training data, and OpenAI respects website owners who block its data-collection bots.
A Glimpse Into the Future
With Google’s Gemini 2.0 Flash also entering the AI image race, GPT-4o’s “all-in-one” approach could shape how we create and interact with content going forward. A future where anyone can simply talk to AI and have it write, draw, and edit is no longer science fiction. It’s happening now.


