DALL-E was launched 4 years ago by OpenAI as an artificial intelligence language model that can generate images. After three years, they came up with DALL-E 3 which remains used in ChatGPT until today when a new model was introduced.
Image generation in ChatGPT now uses the GPT-4o model. GPT-4o has become multi-model with various updates. Each image generated is more accurate and clearer according to the user's wishes, only this time it takes more time to process. The most interesting thing is that this model understands the text deeply so each generation is no longer a nonsense sentence or word.
This model also allows users to edit images. Insert a source image, then insert a logo and can give instructions to generate a new mock image similar to the source image. This model is limited to ChatGPT Pro, Plus, and Teams customers with access to free users postponed to a later date.