It’s a long weekend, and I was bored, so I generated some random AI art to entertain myself. I used ComfyUI and ran some small models on my PC.
You might be wondering why not use ChatGPT / Co-Pilot / Gemini? Here’s why:
- It’s free and there are no queues since you are generating the images on your own hardware.
- You are not limited to specific image sizes or aspect ratios.
- You are not limited to text-to-image prompts. You can generate image-to-image and even videos too.
- You can use a wide variety of models. There are models which are good at generating photo-realistic images, models good at anime style images, etc.
Here are some samples that I generated:
 " | Prompt A young woman with cat ears and glasses sitting in a library Model: SD3.5 |
 | Prompt Miniature of a quaint town in Kyoto Model: SD3.5 |
 | Prompt A knight in shining armor carrying a AK47 facing a black dragon. Bolts of lightning could be seen in the background. Model: SD3.5 |
 | Prompt Long shot. An anthropomorphic rabbit wearing a trench coat walking along a dimly lit street Model: Wan2.1 |
ComfyUI can be intimidating to new users, so I wrote a quickstart guide here: ComfyUI Quickstart