LoRA (Low-Rank Adaptation)
Definition
An efficient way to adapt an existing AI model to a specific style or task without retraining the whole thing. Instead of rebuilding a model from scratch, LoRA adds a small, lightweight layer of adjustments on top. It's popular in image generation, where people use it to teach a model a particular character, art style, or look, cheaply and quickly. A small file does what would otherwise take enormous resources.
Why It Matters
It's part of why a thriving world of custom AI models exists, especially for images. If you've seen communities sharing tailored AI styles, LoRA is often how.
What does that look like in practice?
An artist uses a LoRA to teach an image tool their own painting style from just a handful of examples.
What people actually mean when they say this
When someone shares 'a LoRA' for a look or character, they mean a small add-on that nudges an existing model, not a whole new model.
Last reviewed: June 2026
Related reading
In the AI Lineup
- MidjourneyThe image generator known for sheer beauty. If you want pictures that look like art, this is the one people reach for.
- ChatGPT Image / DALL·EOpenAI's image generator, built right into ChatGPT. Describe a picture in plain words and it appears, no separate tool needed.
- Stable DiffusionThe open image generator that runs on your own machine. Free, private, and endlessly customisable, if you're willing to tinker.
- Adobe FireflyAdobe's image and video generator, built to be commercially safe. Made for people who need to use the results in real work.
Knowing the word is just the start.
We help non-tech people go from looking things up to actually using AI confidently. Tutorials, prompts, plain English. All of it, for less than a cup of coffee a month.
Show Me How This Works →Start free. Cancel anytime. No judgment, ever.