OpenAI brings native image generation to ChatGPT [English]

Author: Matthias Bastian | Source: THE DECODER | Read the full article in English

OpenAI has recently enhanced ChatGPT by adding a new feature that allows users to generate images directly within the platform. This update replaces the previous DALL-E integration and aims to provide users with more reliable results and fewer restrictions on content. The new image generation capability, introduced with the latest version of ChatGPT, is set to be available to all users, from free accounts to enterprise clients, with plans for API access for developers in the near future.

The upgraded system is designed to process both text and images simultaneously, which helps in producing more accurate and contextually relevant images. It can handle multiple objects in a single image while maintaining their relationships, making it particularly effective for creating complex visuals like infographics or logos. Users can also interact with the AI in a conversational manner to refine their image outputs, allowing for a more personalized and iterative creative process.

While the new image generation feature shows promise, OpenAI acknowledges that it still has some limitations, such as occasional inaccuracies in rendering and challenges with non-Latin text. The company is actively working on improvements, including better editing options for generated images. This launch comes in response to similar advancements made by other tech companies, highlighting the growing trend of integrating image generation capabilities into AI platforms.

[Read More (translated)]