OpenAI has upgraded ChatGPT’s image generation with its advanced GPT-4o model, replacing the older DALL-E 3. The new model offers more accurate, detailed images with improved text rendering, photorealism and object consistency. However, it takes longer to generate results and still has limitations, such as issues with cropping, non-Latin text and fine details. OpenAI has also added safety measures, embedding C2PA metadata in images and blocking harmful content. The feature is rolling out to Plus, Pro, Team and Free users, with Enterprise, Edu and API access coming soon.
Source: Android Authority
