Indistinguishable from reality: OpenAI released a new image generation model, GPT Image 2 (10 photos)
Category: Neural network, PEGI 0+
Today, 02:43
OpenAI has released a new image generation model, GPT Image 2, which is a cut above its competitors. Images are now rendered down to the smallest detail, with an emphasis on photorealism and minimal hallucinations. The model also better understands complex scenes.
The most noticeable difference is improved text generation. Text is now truly legible, including complex and even handwritten text. This could simplify the work of designers and interface creators.
The model is already available to users. You can try it out in ChatGPT. 
