Banana.

The new & best AI for images. Yes, it's better than ChatGPT.

The new best AI to generate images. Yes, it's better than ChatGPT.

Google released “nano-banana” right inside their Gemini app.

Here are my favorite case studies & how to access it.

---

How to access “nano-banana”

Three simple steps:

1. Go to Gemini.

2. Make sure you select “2.5 Flash” (top left).

3. Then click on the “tool” icon & “Create images”.

Why is it so good?

It can generate images, sure. But ChatGPT & Midjourney are better.

But what “nano-banana” does better than any other AI: editing images.

Think Photoshop meets ChatGPT.

Here are my favorite case studies:

---

My favorite image-editing prompts

A list of my favorite prompts.

1 - Change your hairstyle.

Prompt: “change my hairstyle to rock-long hair like I'm a rockstar”

2 - Restore old photos.

Prompt: “restore this old photo with today's realistic colors“

3 - Replace X with Y.

Prompt: “change X on right/left with Y”.

4 - Remove the background.

Prompt: “Remove the background here”.

PS: I still prefer Canva for this. Built-in tool is great.

5 - Extract this from this picture.

Prompt: “extract X, only the X”.

6 - Make a 3D model.

Prompt: “Make [X] daytime & isometric. [x] only.”

7 - Change the color of anything.

Prompt: “Change the color of [X] to [Y].”

---

I already have a dozen apps I want to build with this model.

• photoshop for noobies

• photobomb any picture

• hang out with celebrities

• quick 3D model as a service for architects

• change your haircut before the hairdresser

• generate your living room with AI (& buy with Google Lens)

You guys get to decide what I build (and I’ll do it with Blink).

I am creating a conversation on my Substack chat here:

PS: if you enjoyed this newsletter, send it to a friend!

PPS: what would you like to me to write next? leave a comment:

blog archive: https://docs.google.com/document/d/1pWuMCBVQo1zKcgKltX_BZxAr31KgxmOlp3Vzvmc5Hxc/edit?usp=sharing