Published: 2026-09-07
I've been working on a comic generator as a way of embracing AI. I use it daily in my work and the idea behind this comic generator was based on my desire to showcase humor but lacking the skillset to do it. I can draw, but I can't draw well. As a result, I don't enjoy drawing. I can write, but I also don't write well. That's why I have AI write things for me.
But that doesn't mean that AI has to create everything. The comic generator is a way of creating a template for allowing people to be creative. If you can draw, add your drawings. If you can write, write. There is no requirement that you have to let AI do everything but that is also a choice.
The design of this was mostly written by Astra Max, with my design choices. It works against OpenRouter API and generates images in layers - think backgrounds, props, characters. Generate a background for each comic panel, generate the characters, remix the background or characters based on the need. The goal is to eventually publish a proof of concept that allows me to be able to continue creating these comics based on moments I feel are worthy to share.
I know that this is an evil blight on humanity's creativity. I'm aware that it is slopware. I also don't care anymore. It's a fun project and I've enjoyed playing around with it.
Anyway, here are some example comics. There are still consistency issues to be worked out. That is why I am developing in layers which allow you to copy between panels. This will enforce consistency no matter which image model is being used and I do plan to support more than just nano banano, though that is the most performant image model at this time.
But that doesn't mean that AI has to create everything. The comic generator is a way of creating a template for allowing people to be creative. If you can draw, add your drawings. If you can write, write. There is no requirement that you have to let AI do everything but that is also a choice.
The design of this was mostly written by Astra Max, with my design choices. It works against OpenRouter API and generates images in layers - think backgrounds, props, characters. Generate a background for each comic panel, generate the characters, remix the background or characters based on the need. The goal is to eventually publish a proof of concept that allows me to be able to continue creating these comics based on moments I feel are worthy to share.
I know that this is an evil blight on humanity's creativity. I'm aware that it is slopware. I also don't care anymore. It's a fun project and I've enjoyed playing around with it.
Anyway, here are some example comics. There are still consistency issues to be worked out. That is why I am developing in layers which allow you to copy between panels. This will enforce consistency no matter which image model is being used and I do plan to support more than just nano banano, though that is the most performant image model at this time.
.png)
.png)