Thorough Comparison of NanoBanana2 and Pro | Recommended Tools Too! [Image Generation AI / AI Illustration / AI Art / AI Utilization / GoogleGemini]

Good morning. This is Kinu.
Thank you for coming to visit.

Google's image generation model 'NanoBanana2 (Gemini 3.1 Flash Image)' has been released, so I tried it out right away.

In this article, I will explain a comparison and verification of the conventional 'NanoBananaPro' and 'NanoBanana2' as well as how to use them.

Please take a look at how to utilize the latest model to make creating materials and banners even easier✨

◆ Features of NanoBanana2 | Dramatic Evolution in Japanese Understanding and Composition Sense

While NanoBanana2 is based on image generation from text,
with this update, 'Japanese language comprehension' and 'sense of composition' have improved significantly.

① Image generation from text (basic function)
It generates high-quality visuals from scratch just by providing a blueprint called a prompt.
In particular, the stability of Japanese text has increased, making it easier to use for creating materials and banners.

② Editing with image + text (inpainting)
This is a function that allows you to specify instructions in words for the finished image, such as 'fix just this part' or 'port this style'.

③ Style transfer from multiple images
This is an advanced combination technique, such as 'drawing the composition of B using the art style (worldview) of image A.' Since you can copy your favorite textures to other images, it is very useful when you want to create many images with the same taste, such as for videos or illustrations for notes.

◆ How to use Nano Banana 2 and Nano Banana Pro

The answer to the question 'Which one should I use?' is very simple.

Nano Banana 2: Rapid generation, accurate adherence to instructions, search integration
Ideal for creating rough banners, diagrams, and visuals for materials.

Nano Banana Pro: Maximum factual accuracy, high-definition depiction.
Suited for tasks that require the pursuit of texture, final finishing, and high quality.

◆ How to use Nano Banana2 and why it is recommended

Nano Banana allows image generation from several AI tools, but personally, I recommend generating images with Google Flow and Google Gemini.

① Why is Google Flow recommended?

• Try the latest model for free: Within specific usage limits, you can experience the latest Nano Banana 2 for free.

• Intuitive interface: Since you can generate while switching aspect ratios, the production flow is completed smoothly.

• Capable of generating 4 images at once

• No watermark

② How to use Google Flow

Access Google Flow

1. Log in with your Google account

2. Click on "New Project"

3. Select aspect ratio "NanoBanana2" or "NanoBananaPro"

4. Enter a prompt and generate

3. Why is Google Gemini recommended?

• The UI is overwhelmingly easy to use (the barrier to operating the tool is low).

• It remembers the conversation context:The AI understands and reflects the content we were just discussing (the brainstorming content) as background information.

Example: After turning your article or text into a manga, if you give the instruction, "I want you to generate a manga in the style of an RPG game screen," then...

Like this!? (lol) It generates a manga with a different atmosphere while carrying over the context from before.

4. How to use Google Gemini

1. Open a new chat screen in Gemini

2. Select "Create Image"

3. Select "Fast Mode" or "Thinking Mode"

4. Send the prompt to generate

[Prompt Collection for Nano Banana]
I am introducing prompts that you can use by copying and pasting.
First, try it out for free, and please use it while referring to the prompt collection✨

◆ Impressions after actually using it

What surprised me was the stability of the Japanese text.
I generated an infographic with the same prompt and compared it with NanoBananaPro, but...
Honestly, I felt that for document creation, NanoBanana2 provides more than enough perfection.

On the other hand, there were times when I felt that NanoBananaPro had higher accuracy in image generation.
For example, a tendency to break down when increasing the complexity of the composition, motifs, or the amount of information in elements became apparent.

Specifically, there were cases where characters would split into two, or hands would have three arms.

Also, I felt a bit uneasy about copyright-related issues, as images were sometimes generated even without specifying a particular character.

That said, I felt that for document creation and banner production, Nano Banana 2 provides more than enough perfection.

While Pro is recommended for situations where you seek quality and high standards, my impression is that it can handle almost all daily production tasks.

◆ "NanoBanana2" Creative Gallery

[Manga Edition]

It looks good... there don't seem to be any major problems.

[Advertising Banner Edition]

If you craft the prompt thoroughly, Nano Banana 2 also produced sufficiently good output!

[Infographic Edition]
*The text in the document is placeholder text.

Here too, the text was stable, and there were few failures in the placement of figures and motifs.

◆ Summary | How much can you do with NanoBanana2?

As a conclusion at this point,

• For document creation and banner production, you can almost entirely rely on NanoBanana2

• Pro has the advantage in texture and final finishing

• We recommend trying it for free via Google Flow
  is the impression I have.

Moving forward, I plan to continue sharing verification results and exploring the possibilities of NanoBanana2 together with you all.

First, try it out for free, and while referring to the prompt collection,
please try applying it to your own creative work!

Thank you for always watching.
I will continue to deliver experiments and verification results!
I hope you have a wonderful day today✨