How LLMs Work102 · Module V · Lesson 29 of 31
Article · 12 min

Quantization, intro

Running the same model at a fraction of the cost.

Summary

Quantization stores and computes model weights in fewer bits — 8, 4, even 2 — instead of 16 or 32. A 70B model in fp16 needs 140GB; in 4-bit, 35GB. Modern quantization loses little quality on most tasks and makes local deployment of frontier-scale models tractable.

Objectives
  • 01Explain what quantization compresses.
  • 02State typical quality/size tradeoffs.
  • 03Distinguish weight-only from activation quantization.
The Lesson

The compression

Store each weight as a low-bit integer with a shared scale, instead of a 16-bit float. GGUF (llama.cpp), AWQ, and GPTQ are the three dominant formats. 8-bit is nearly lossless; 4-bit costs a few points on hard benchmarks; 2-bit degrades noticeably.

Weight vs. activation

Weight-only quantization shrinks storage and reduces memory bandwidth (the bottleneck for most inference). Activation quantization additionally speeds up computation on hardware that supports low-precision math. Most local runtimes quantize weights only.

Key Ideas
  • Quantization trades a little quality for a lot of size.
  • 4-bit is the standard for local deployment as of 2026.