Google will let searchers create images without leaving AI Overviews, turning a written prompt into a finished visual on the results page itself. The company is powering the feature with its Nano Banana image model, according to Search Engine Land. The logic is simple. Every capability added to AI Overviews gives a user one less reason to click away, and image generation is the latest one folded in.

Google frames the tool as turning “a simple text prompt” into a finished graphic built entirely from scratch, not pulled from an index of existing photos. That is a real distinction. Traditional Google Image Search surfaces pictures that already exist somewhere on the web. Nano Banana instead renders something new, inside the same answer box that already summarizes text queries.

The stakes extend beyond blue links. Image-generation sites, stock-photo libraries, and AI art tools have long absorbed the demand that search used to route toward, whenever a user wanted a visual that did not exist yet rather than one that did. For many of those tools, search is the primary discovery channel: a user types “make an image of” into Google and clicks through to whichever generator ranks. Folding that capability into the Overview intercepts the query before the click, the same way AI Overviews already intercept a factual question before it reaches a publisher.

AI Overviews already changed click-through behavior for the informational queries where they appear most, a trend search marketers have tracked since Google’s earlier rollouts of the feature. Extending that same summarization logic to visual output follows the same mechanism: answer the need before the user leaves the page.

Google has not said how much of Google Image Search traffic, or how much third-party image-generator traffic, this feature is expected to absorb. That gap matters. It is the one number that would tell publishers and image-tool makers how large this shift really is.

The feature is rolling out over the coming weeks in English, limited to regions that already support image creation in AI Mode, Google’s conversational search experience, Search Engine Land reported. Google paired the launch with a redesign of Google Image Search, timed to that product’s 25th anniversary. Both updates point the same direction. Keep image-related search activity, whether finding a picture or making one, inside Google’s own pages.

Search and content teams should isolate which image-heavy queries already trigger AI Overviews, since those are the queries most likely to lose referral traffic first. Anyone whose business depends on people searching to create a visual, not just find one, should treat this rollout as a traffic forecast to run now, not a future risk to watch.

Search Engine Land, in reporting by contributing editor Barry Schwartz published July 14, 2026, first detailed Google’s rollout of image generation inside AI Overviews.