Parameters
Definition
The internal settings a model adjusts as it learns, and the rough measure of its size. Modern models have billions of them. Loosely, you can think of parameters as the dials the system tunes during training to get better at its task. More parameters can mean a more capable model, though it isn't a simple case of bigger being better, and it comes at a cost in computing power.
Why It Matters
When you see a model described as having "70 billion parameters," this is what that means. It's a size label, useful for comparison but not the whole story.
What does that look like in practice?
A model billed as '70 billion parameters' has that many internal dials it tuned while learning.
What people actually mean when they say this
When someone boasts about 'billions of parameters,' they mean the model is large, which can help, but isn't the whole story of how good it is.
Last reviewed: June 2026
Related reading
From the Unspun Blog
- U.S. data centers could account for about 20% of national electricity consumption by 2035 — and that already touches bills on some gridsBNEF projects U.S. data centers could account for about 20% of national electricity consumption by 2035. What that means for bills and regional grids.
- Why Does AI Make Things Up, and Can I Actually Trust It?AI tools make things up. It's called hallucination, and it's not a glitch. Here's what's actually happening and which tasks are genuinely safe.
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 $7 a month, and the first seven days are free.
Show Me How This Works →Start free. Cancel anytime. No judgment, ever.