PKAI Art Course
Generate NSFW AI Art.
No Censorship. No Monthly Fees.
My guide to local AI art generation using free and open source models like Stable Diffusion running on free software, ComfyUI.
One-time purchase · Self-paced · Lifetime Course Access· 16 modules
What You'll Be Creating
What You Get
Lifetime access to training course · My full ComfyUI workflow · 1-on-1 Discord support · 2 set requests · 1 month of PKAI content access · Mega archive of published sets
Why I Prefer Local Generation
Online services limit what you can generate via content policies. Monthly subscriptions add up. Running open source models locally means unlimited generation, full control, zero censorship — and it can be faster than any online tool.
No Censorship
No Monthly Fees
Unlimited Output
16 Modules. Continuous Development.
Everything You Need.
From your first image to professional-quality NSFW sets — at your own pace.
Foundation Track
🚀 Module 0 — Welcome & Course Roadmap
⚙️ Module 1 — Setup ComfyUI
🛍️ Module 2 — CivitAI: Your Model Hub
🎨 Module 3 — Your First Generations
💬 Module 4 — Prompting Mastery
🔧 Module 5 — ComfyUI Foundations
🧩 Module 6 — LoRAs: The Complete Guide
📐 Module 7 — Upscaling
🖼️ Module 8 — Img2Img & Inpainting
Advanced Track
👁️ Module 9 — Detailers
🎛️ Module 10 — ControlNet
🎲 Module 11 — Dynamic Prompts & Wildcards
🗺️ Module 12 — Regional Prompting
✨ Module 13 — Frontier Models
🎬 Module 14 — Animation with Wan 2.2
🏋️ Module 15 — Training Your Own LoRAs
⭐ Module 16 — Building Your Style & Going Pro
What can you expect on your PC?
Find out how many images per minute your GPU generates at standard resolution of (1216×832 SDXL).
YOUR INSTRUCTOR
Hi, I’m PKAI.
I’m an impatient artist and lifelong tech enthusiast. I’ve been generating AI art since early 2025. I’m just addicted to it and this business is my healthy outlet to channel it.
You don’t need to take my course to get into this hobby. This course is here if you want to gain from my experience and have direct contact and input from me as you learn. I think you will love the way I structure the course and I’m sure you will find my enthusiasm infectious.
I’ll spoil it for you if you want to dive in on your own. The software I use is ComfyUI, The models I use for images are tons of different Illustrious and NAI based models from CivitAI. The model I use for video right now is wan2.2.
I also just really love AI. I’m sorry; I know it’s taboo to say what with the world being destroyed by it and RAM costing 500% what it should.. I just can’t make myself hate it.. It’s just too cool.
What Students Say
“I just gotta say, the way you’ve been pulling back the layers one at a time has been perfect. I think I’m ready to learn about controlnet.”
— My first student (T^T)
“i’ve already had so much fun with the basic stuff, and this is going to give me fun for months, or maybe even years to come“
— Course student
“I feel like I had been previously using a limited demo version of comfyui after all the stuff you’ve taught me.”
— Course student
Get Full Access
One-Time Purchase
$120
Lifetime course access · Self-paced · All modules
✓ Full course access — all 16 modules
✓ ComfyUI workflow download
✓ Regular content updates
✓ 2 set requests
Questions?
What do I get with this subscription?
- Lifetime access to training course, meaning you will have access to updates I make to the course
- My full ComfyUI workflow
- 1-on-1 Discord support with me
Additionally, I’m including all the content of the Balance tier in this tier. That comes with:
- 2 set requests with at least 200 images and 20 videos per set.
- 1 month of PKAI published content access
- Mega archive of all published sets
Do I need a powerful PC?
No. You can rent cloud machines for as little as $1/hour if you don’t have the right hardware.
You would be surprised what you can accomplish with local models on your machine though. An Nvidia GPU from after 2020 with at least 8GB VRAM will probably give you an enjoyable experience. Please scroll up and use the calculator I built off real benchmark data to learn more about what you experience will be. This calculator can also help you if you are thinking about upgrading.
What software is covered in this course?
I use ComfyUI, the gold standard in local AI. That primarily what this course will cover. Even if you rent a cloud machine for super cheap, you will still be running ComfyUI. I’ll be adding alternatives to ComfyUI in the future.
What models are covered in this course?
This course covers some of the best free and open source image and video generation models out there. The brunt of the course is focsed on Stable Diffusion based models. We also cover Wan 2.2 for video generation and other frontier models such as Flux, Qwen, and Z-Image. I follow the space as a hobby and will keep this course up to date as developments occur.
What 'styles' are covered in this course?
I will give you the tools you need to essentially replicate any style you want. I’m also available to give you direct support and guidance.
How long does the course take?
It’s fully self-paced and there’s no expiry on your access. You will be generating images of your favoriate characters shortly after installing the software. The early dopamine hit really helps to carry you through the course. Most students work through the course over 4–6 weeks. The concepts build on each other and you will want to give youself time to impliment concepts in practice before moving to the next one.
I have no art experience; is this course for me?
AI is a tool that can help you get the ideas in your head onto your screen. You do not need any experience or a background in art to have a fun and satisfying experience making content.
If you have some art experience, an eye for colors or composition or just a lot of creativity, you will find that this translates very well to generating AI art.
You will find the speed at which you can experiment and iterate to be very satisfying.