How to Use Multi-Image Prompting in Nano Banana

In this video, we are going to see how we can do the multi-image prompting with nano banana. So, basically on our left,

nano banana. So, basically on our left, we have combined many objects and a picture of myself. I've added glasses,

some belts, watch, shirt, a briefcase, and a car. So, basically what we do is just open Canva or Photoshop. on the

blank canvas. Just paste all of the things you like and then you will give a single command to nano banana and what this will do is this will combine all of

the things you have mentioned on the board and bring this to life. Look at

this guys. He's wearing the same glasses. This come from my original

glasses. This come from my original outfit with this shirt in the background holding the same suitcase next to the same car and he's wearing the same watch

and a belt and all of the stuff like that. So basically what you want to do

that. So basically what you want to do is let's say that you want to try different outfits. You can do whatever

different outfits. You can do whatever you want. You can just get the perfect

you want. You can just get the perfect combination of the outfits from Pinterest or Instagram and you can use

that images next to yourself and just give a simple command and you as a model will wear all of this stuff. So the

command is very simple and straightforward. What I'll do is I'll

straightforward. What I'll do is I'll just link this with the resources section of this video. But but just remember that this will not work directly with Gemini. What you need to

do is you need to uh use the open art which uses nano banana as a third party software and there you can easily generate that. So although it's

generate that. So although it's nanobanana but if you go to gemini directly and we paste this command a confident young man in a stylish casual

smart outfit standing beside a sleek black supercar something like that. So

what this will do is this will generate the results like this. It has not generated perfect results. It has kept the other images or other objects

separate. So we need to use open art for

separate. So we need to use open art for this. Just type open art on Google

this. Just type open art on Google search and you will get to this area where we have to make a free subscription and you get some credits

from this. Here we just go to the image

from this. Here we just go to the image creation section. We just define our

creation section. We just define our custom settings is set to none. The

model or character should be defined by nano banana. Now I like this personally

nano banana. Now I like this personally because it supports all of the other features like V3 and all GPT images and

it uses some other stuff as well. So

let's use nano banana for the reference.

The prompt I'll just simply paste this prompt and the reference we will use the reference image like this one. And let's

say we only want to generate one image because I have limited credits available even though I have 10 credits. I need 15 credits to generate this. So I've

already generated this. After that what you need to do is just click on create and it will generate the um finalized images for you. Something like that. So

if you like this you can upgrade this to 4,000 credits per month which is only $7 per month. But that is very interesting

per month. But that is very interesting to do and you cannot do this directly in Gemini.

But if you try Lamarina in the first try with assistant A, this is completely a different person. But in the second

different person. But in the second image, it quite looks like myself. So

you can use Ella Marina for that. In

some cases, it works better than Gemini.

And also what you can do is just try defining a character name. Let me just define the character. I'll just put the name here. And now let's start a new

name here. And now let's start a new chat. Upload the image. This is the one.

chat. Upload the image. This is the one.

Now instead of just putting a random command, let's use the name as a reference. Sad

is wearing a stylish casual smart outfit like this. And let's see what changes we

like this. And let's see what changes we get in the results. Now the benefit of using names is it will just continue the reference even though it we define it in

a different scene like sad is standing next to a theme park or something like that. No, but it hasn't done a good job

that. No, but it hasn't done a good job here. So you have to do a couple of

here. So you have to do a couple of tries. In the first try it worked best.

tries. In the first try it worked best.

So I believe the general command is working better here. Let's try

generating it again to just confirm this. and in the second try it is not

this. and in the second try it is not producing that much of good results. So

if that happens what you can do is just close the applications or just close the browser altogether. In some cases you

browser altogether. In some cases you might even need to restart your PC and then try giving the same command again and it will generate much better results. So that's about it. Just try

results. So that's about it. Just try

this by yourself and I will see you in the next