OpenAI has launched a new feature called Images in ChatGPT. With this feature, users can create artistic and attractive images directly inside ChatGPT using the GPT-4o model.
Available for All Users
According to the company, this feature will be available for all versions — ChatGPT Plus, Pro, Team, and Free. Free users can create images within the same limit set by DALL·E. Earlier, free users were allowed to generate three images per day using DALL·E 3.
More Powerful and Improved Model
OpenAI researcher Gabriel Goh explained that this new system is much better than older models. It uses a new technology called Binding, which helps the model recognize colors and shapes more accurately. While older AI tools often struggled to separate different objects, this new model can identify 15 to 20 objects clearly in one image.
Perfect for Infographics and Posters
The company says the system can easily create infographics, posters, stickers, and scientific illustrations.
Better Text in Images
The new system also improves text rendering inside images. Now, the letters and words in the images will appear clear and readable. Previously, many AI tools could not render text properly. Gabriel Goh said the team worked for months to improve these features.
Focus on Safety and Ethics
OpenAI has also added safety measures to this system. The company says it has taken steps to stop the creation of harmful or misleading images. All images created by OpenAI tools will have C2PA metadata, showing that the image was generated by OpenAI.
