Creating a Manga Character with Anifusion: My Prompting Adventure

As someone who’s passionate about innovation and curious about the intersection of creativity and technology, I recently decided to test out Anifusion, an AI-powered manga creation tool that promises to turn text prompts into visually compelling manga panels. The idea of generating characters and scenes with just a few words was too tempting to pass up.

So I dove in, with excitement, curiosity, and a healthy dose of skepticism. What followed was a journey of trial, error, refinement, and a few unexpected surprises. If you’re thinking about using Anifusion or any AI art tool to create manga characters, here’s a behind-the-scenes look at what it’s really like.

Step 1: Getting Started

The interface is clean and intuitive. I clicked New Canvas, hovered over the right-hand box, and selected Create Manga.

From there, I tapped the Prompt button at the top of the right panel, and the real fun began.

I had a specific character in mind: A samurai girl with dark, curly black hair, a button nose, and piercing dark blue eyes. She’s holding a katana and fighting off an evil gremlin. It was a very specific and vivid image I had in my head with intense features. I wanted to see how close Anifusion could get.

Prompt #1: Long and Specific

My first prompt was detailed and descriptive. I included everything from her physical features to the setting and action. I figured the more information I gave the AI, the better the result would be. So I clicked on the generate button.

Result?
Close... but not quite. The character looked fierce, and the composition was dramatic but the AI gave her glaring red gremlin eyes instead of the piercing blue ones I had asked for. It was as if the gremlin’s traits had bled into my heroine. A little unsettling, a little funny but not quite what I had envisioned.

Generated by anifusion.ai

Prompt #2: Refining the Details

I went back to the drawing board, literally. I refined the prompt, emphasizing the girl's features and removing the gremlin entirely.

Result?
The hair and sword were spot-on this time. Progress! But now her eyes weren’t visible at all. They are visible but not clear and you can barely make out the colour. It felt like the AI had dodged the challenge of rendering them.

Generated by anifusion.ai

Prompt #3: Getting Closer

I adjusted the prompt again, this time focusing more on her facial features and expression. I wanted those eyes to shine through.

Result?
Finally, the blue eyes appeared! But they weren’t quite as piercing or vibrant as I had imagined. They were there, yes but it just was not what I’d imagined. The intensity I was aiming for just wasn’t coming through and the image had been changed entirely, so she no longer had the samurai sword.

Generated by anifusion.ai

Prompt #4: Shorter and Sharper

In this prompt I kept it short again but tried using capital letters to emphasize the characteristics more and emphasized with words like ‘very’ and ‘big’. I thought maybe the AI was getting overwhelmed by too much detail. I stripped the prompt down to its essentials, focusing tightly on just the eyes.

Result?
Much Better! Definitely closer to what I had imagined, at least the eyes were anyway. The eyes were a much darker and appealing colour of blue and the composition was cleaner. But it still was not the entire character I had envisioned, she was no longer a samurai and other characteristics had been enhanced, which I had not asked to be.

Generated by anifusion.ai

Final Touch: Enhance Face

Then I noticed a feature I hadn’t tried yet: Enhance Face. Intrigued to see what it could do, I clicked on the image, scrolled to the Image Effects section in the right-hand panel, and activated it.

Result?
I could see that the AI had made the effort to enhance the eyes but it was only a slight adjustment and I was a bit disappointed with the results. I could have possibly had a bit too high expectations but I was expecting a bit more of a transformation.

Generated by anifusion.ai

What I Learned

Prompt Crafting Is an Art
The way you phrase your prompt matters. Small changes in wording can lead to dramatically different results. Being specific helps, but sometimes less is more.
AI Has Its Own Interpretation
Even with detailed prompts, the AI might interpret things differently than you expect. It’s not a mind reader, it’s more of a pattern matcher. Sometimes it blends elements in unexpected ways.
Iteration Is Key
You won’t get the perfect image on the first try. It takes multiple attempts, refinements, and tweaks to get close to your vision. Patience is part of the process.
Built-In Tools Can Help
Features like “Enhance Face” could make a difference. Don’t overlook the editing options available within the platform, they might be the key to unlocking the final polish.

Final Verdict

Anifusion is a powerful and promising tool. The image quality is impressive, and the interface is user-friendly. But like many AI generators, it’s not quite ready to replace human intuition and artistic nuance. It’s a great companion for brainstorming, prototyping, and experimenting but if you’re looking for pixel-perfect precision, you may still need to bring your own artistic skills to the table.

That said, I had a blast exploring it. And I’ll definitely be using it again, maybe for character concept art, storyboarding, or even just for fun.