Microsoft has announced that Bing Image Creator now offers the latest GPT-4o image generation model from OpenAI for free, providing users with a more powerful AI image creation experience.
Significant Advantages of the New Model
Earlier this year, OpenAI integrated image generation capabilities directly into GPT-4o, eliminating reliance on independent models such as DALL-E. Compared to DALL-E3, GPT-4o performs more accurately in text rendering, and can accurately understand user intent using its extensive knowledge base and conversational context.
The new model also supports image editing features, allowing users to upload existing images for modification or use them as creative references to generate entirely new visual effects. However, since GPT-4o can create more detailed images, the rendering time is accordingly extended.
Two-Model Selection Strategy
Based on user feedback, Microsoft has decided to retain both DALL-E3 and GPT-4o models, allowing users to freely choose when creating images. The Bing team said: "We know that without freedom, creativity has no meaning. Now, simply click once to switch the model to DALL-E3, quickly generate AI images, and get more options."
Usage Policies and Access Methods
Bing Image Creator maintains a free usage policy, offering 15 fast generation opportunities. After exceeding the limit, each image requires 10 Microsoft Rewards points for fast generation, but users can still create images unlimitedly at standard speed.