Starting Thursday, Google's Search Generation Experience (SGE) will feature a new ability to create images directly from text prompts, the company announced.

Microsoft has been offering images for Bing chats since March of this year, based on OpenAI's DALL-E model.
Now, the user can type the relevant query into the Google search bar, and when the input is complete, SGE can create some images according to the prompts, and the user can choose from them.
Google spokesman Craig Ewer said the tool is powered by the Imagen family of artificial intelligence models.
In a GIF, Google shows how the tool creates images based on the prompt to "draw a capybara wearing a chef's hat and making breakfast."
And as the GIF shows, users can also select the desired image and then edit and optimize the input text information.
In addition, users can also create Images generated by Google artificial intelligence with Google Images. When you browse through Google Images search results, you may see a prompt box prompting you to generate an image based on your query - the resulting image will be displayed in the sidebar.
Hema Budaraju, a product manager at Google, stressed in the post that the company is working to ensure that its image generation tool is rolled out in a responsible manner and that the tool is not allowed to generate images that violate the company's relevant policies.
According to Budaraju, every image made with SGE will have "metadata labels and embedded watermarks to indicate that it was created by artificial intelligence"; And Google won't let users create images of celebrities.
Google limits image generation to users 18 or older, which means teens won't be able to use it.