AI Tool Guides
Best for
Fully customisable open-source image AI
Works on
Local (Python), cloud
Alternatives
FLUX, Midjourney
Watch out
Complex setup for local use; easier via services like RunDiffusion
What It Does
Stable Diffusion is the open-source image generation model that started the AI art revolution. Because the model weights are freely available, it can run on your own computer with no internet connection, no usage limits, and no content filters. The ecosystem of community-built tools, fine-tunes, and extensions is enormous.
Setup in 5 Minutes
, install the UI, point it
at the weights, and start generating.
- If you lack a GPU, use the free tier on
huggingface.co/spaces to run it in the cloud.
Try This
Install ComfyUI and load the SDXL model. Generate an image with:
"A cosy study filled with books and warm lamp light, oil
painting style". Now install a community LoRA (a small fine-tuned
adapter)—search CivitAI for a watercolour LoRA—and regenerate with
the same prompt. The stylistic shift demonstrates how fine-tuning
works at a practical level without any training on your part.
Follow Along
Follow Along — Style Transfer Experiment:
Generate it with the default SDXL model.
- Download five different LoRA style adapters from CivitAI (e.g. watercolour, pixel art, anime, oil painting, and pencil sketch).
- Generate the same prompt with each LoRA applied, keeping all other settings identical.
- Arrange the five outputs plus the base image in a 2\(\)3 comparison grid. Label each with the LoRA name. This is a hands-on lesson in how fine-tuning reshapes a model's output.
More in Creative Tools
30DALL-E (via ChatGPT)— Image generation inside ChatGPT conversations
32Midjourney— Highly stylised, aesthetic AI art
36Leonardo.ai— Game art and character design generation
Free tier38Udio— Music with precise genre and style control
Free tier40Soundraw— Royalty-free background music for videos
42Pika— Quick short video clips from text
Free tierAI Analysis
Frameworks from the aiborg Handbook — powered by Claude