Now we are accustomed to general AI to Create images From text hints. These tools continue to become more complex and universal – even if many questions remain nearby Using energyCopyright violation and long -term influences.
The last major update for deployment in this category Ay For the Google Gemini application. He is known as Nano Banana after he had during testing – in private, it is called Flash – the Gemini 2.5.
In addition to general improvements, it is really excellent in editing images, where generative AI previously fought. Editing will not be consistent, or will change the incorrect parts of the images, or will not follow the instructions properly.
“This update allows you to mix several images with one image, maintain the consistency of symbols for a rich story of stories, make targeted transformations using a natural language and use the global knowledge of Gemini to create and edit images,” – Says GoogleField
Using the Gemini 2.5 Flash image, you can be much more specific and accurate in your modifications, so it is almost how to have instant access to the Photoshop expert. We have collected some of the best ways to use it in the application of twins on the InternetField to start, click the icon of the instrument slider in the entry field, then Create imagesAnd then click + (plus) button to choose the original image.
As always, everything that you enter into the twins can be used For further training Google AI model And sometimes you can see human reviewers – so avoid sharing something sensitive or confidential. To stop this, you can tell Google not to save any of your chats or images in Gemini, clicking Activity On the left navigation panel.
Make accurate changes
Sometimes you just want to change one specific part of the photograph: maybe you want the car color to be blue, not red, or you need a lamp remote from a shot – or you want to add a lamp. Here, twins can get quick and effective results without the need for a full image editor.
Suppose you have a photograph of how someone does something, but you are not completely satisfied with how they look, although you want to save most of the untouched scene. As soon as your image is downloaded, use the hint, for example, “change its jumper color to red” or “give it curly blond hair”, and the twins create a way out for you.
AI does not always give you perfect results, but most of the time works well. You should see the requested editing, and the rest of the picture remained unchanged. If you do not get exactly what you are looking for to start, you can try again with a slightly different hint (although follow The limits of use) – Try and be as accurate as possible in your requests.
Think about some other changes for which you can turn to the correct image editing tool, and twins will usually be able to remove them if you describe them well enough. For example, you can ask you to blur the background for the main item of the image or bring it into a shot in a shot (maybe some kind of dust or scratch).
Consistently combine and manipulate images
Something else, on which the Flash model of the Gemini 2.5 is excellent, this removes the image or several images and converting them in several ways. If you have a self -portrait, you can put yourself in a completely new place or give yourself a completely new outfit (or both) – AI should be smart enough to make sure that the picture still looks like you.
Try to upload your photo, and then ask the twins to change the background behind you: maybe you want to be on the beach, on the field or light up. You can also save the same thing in the picture, but to adjust the time of day or the weather that does not look a completely new scene created by AI.
The association of images also works very well – just choose two or more paintings when you press + (plus) button inside the box. Give the twins a photo of you and a photo of the clothes that you found on the Internet, and you can request an image on which you wear this clothes. Or, if you have separate photos of your dog and your cat, you can collect them together.
The combination of such images may require a little more trial and error and thoroughly, so you may need to be more accurate with respect to how the images are compiled together. Nevertheless, you should see how the twins support the sequence in several images, so your dog and a cat (for example) will always look like your pets.
Add or remove from the scene
Since you become more confident in what changes that you can get, you can make, you can become more ambitious in your requests. Of course, one of the most common tasks performed in the image editor is to remove objects from the scene, whether it is a random stranger that interferes with your group selfie, or a homeless clothing item that makes your room look dirty.
In order to get rid of anything in your picture, be specifically about what you want to take: tell Gemini “remove the person on the right and fill the back” or “wipe the red jumper that is on the floor”. As long as the image is enough for the twins to be able to fill in the gaps, you must see that the configuration is successfully done.
Of course, the twins can still create people or objects generated by AI, as well as amending, but you do not need to create a whole image from zero, you can add to what you already have. Maybe you would like to see what the book shelf will look like in your office, or you want to fill the street scene with a large number of cars or people.
Gemini will accept hints such as “place a cream sofa along the rear wall” or “add a yellow bus in the background” if you are adding additions. It depends more on the training of the twins, so you may need several passes, but keep in mind that you can always request additional changes or return to the original image and start again.
More transactions, reviews and purchase guides
