Text to Image

Nano Banana 2.1

The new Google image model and the successor to Nano Banana 2. Writes text without typos, lays out infographics and product cards, and keeps faces and objects consistent across up to 14 references. Up to 4K.

20-40 сек

Text to Image

Image to Image

Try See examples ↓

Nano Banana 2.1 Nano Banana 2.1 generation example

"Gift box of the Shalfey ceramics brand, lettering on the box and tag"

From a phone snapshot to a catalog shot

White leather sneakers on a kitchen table under yellow light, a phone snapshot — Before

The same sneakers on a seamless sand background in studio light, one floating above the other — After

The "before" frame is also a Nano Banana 2.1 generation, styled as a seller's phone photo in a kitchen. One request turned it into a catalog shot: the background, crumbs and clutter are gone, the sneakers kept their shape, laces and stitching.

Nano Banana 2.1 generation examples

Real results created by aiTurbo users

"A marketplace product card: headline, three callouts and a badge, all in one request"

"Two references in one frame: the ceramicist and the mug from the product card"

"A four-panel comic: the same heroine in every panel, speech bubbles without typos"

"A pu-erh brewing infographic: five steps, icons and a clean grid"

What Nano Banana 2.1 can do

Text without typos

Product cards, banners, infographics and comic speech bubbles: letters come out crisp and spelled the way you wrote them, Cyrillic included.

Characters that stay themselves

Up to 14 reference images per request. Google states the model holds up to 4 characters and 10 objects: the same face and the same product from scene to scene.