OpenAI has released GPT-4O, now you can create images and edit them and represent text correctly

The Artificial intelligence Keep advancing, we have recently seen how Depseek has flooded news Internet With its optimized model, offering better prices and a series of Differences With the great acquaintances so far. But OpenAI He has not wanted to be left behind and has presented his Image generation with GPT-4Owhich includes Best text representation and Instruction monitoring by the user to improve and modify the images created.

One of the great novelties is the Text improvementthe great failure in the generation of images of the vast majority of models (if not all). GPT-4O is capable of Generate images with legible texts Beyond the scribbles and meaningless symbols that sneaked in most of the time when creating the images. It also has a improved system when modifying and refining the generation of images, after the initial prompt with the description of the image, we can tell the model that go making changes to the already generated modelinstead of modifying the initial prompt and therefore the initial image.

You can also Add your own photos and ask the model to Modify them As we have described above, even adding or modifying other parameters In some of the previous creations. Although there is improvement in this new model, Openai has also discovered Some failures As cuts in the lower images or the difficulty of working with more than 10-20 objects, although it is capable of doing so. You can now Try it from your website.