Google (GOOGL.US) and Musk press OpenAI to upgrade ChatGPT image function: generation speed increased by 4 times, and editing accuracy was significantly optimized

The Zhitong Finance App learned that in fierce competition with companies such as Google (GOOGL.US), AI giant OpenAI continues to strengthen its flagship chatbot products. The company announced on Tuesday that it will launch a major update to ChatGPT's image generation function, which will not only improve the accuracy and editing flexibility of image generation, and increase the generation speed by up to 4 times compared to the previous model, but will also add independent image generation portals to mobile apps and web terminals, which can be used directly without chat interaction.

As an important part of OpenAI's strategy to build an “omnipotent application,” ChatGPT is gradually integrating multiple functions such as search engines, voice assistants, and image and video generation. However, competition in the industry is currently heating up: Google recently launched Gemini 3, a major AI model, and user stickiness of its popular image generation tool Nano Banana continues to increase; Grok, a chatbot launched by XAI company founded by Elon Musk, has also launched a similar image generation function, forming a trend of siege and blocking ChatGPT.

According to previous reports, due to the impact of the Gemini 3 release, OpenAI CEO Sam Altman announced a “red alert” status in an internal memorandum, requesting the team to make every effort to upgrade the ChatGPT function. Just last week, OpenAI just released a more advanced AI model, which focuses on strengthening the application capabilities of ChatGPT in programming, scientific computing, and various office scenarios.

This image function upgrade aims to meet both individual creative and professional production needs. The upgraded system supports multi-dimensional editing of uploaded images: for example, converting the photorealistic style to a watercolor texture, or adding new accessories to the characters in the picture, and all editing operations can preserve the integrity of the details of the original image. OpenAI specifically emphasized that the new model can stably maintain the consistency of lighting effects, character appearance, and scene composition during multiple edits — a pain point that caused previous image models to frequently experience detailed distortion issues in multiple rounds of editing.

Furthermore, the new model has achieved a breakthrough in text generation capabilities, supporting the creation of image content containing more small text, such as generating detailed charts or multi-paragraph text according to user prompts; at the same time, it also performs better in generating multiple clear small-sized face images in a single image.