parameters
The numbers inside the model, learned during training and frozen afterwards. A 27B model has 27 billion of them. Parameter count sets the file size directly: at 2 bytes each, 27 billion is 54 GB, and at 4-bit it is closer to 16 GB. It is the first number in almost every model name.
Weights and parameters are the same thing in practice. Each one is a single number, and running the model means multiplying your input through all of them, or through the fraction of them a mixture-of-experts router picks.
Parameter count is a rough proxy for how much the model knows, and a poor proxy for how smart it is. A 2026 8B model beats a 2023 70B on most tasks, because how long a model was trained and what it was trained on matter as much as its size.