Q4_K_S GGUF models
4-bit k-quant, small. Slightly smaller and a little lower quality than Q4_K_M. 104639 models have a Q4_K_S file. Sizes run from 1M to 1030B parameters. The longest trained context is 4194304K tokens.
What is Q4_K_S?
4-bit k-quant, small. Slightly smaller and a little lower quality than Q4_K_M. On average it stores about 4.6 bits per weight, so a Q4_K_S file is roughly 29% of the size of the same model in 16-bit. The usual sweet spot between file size and quality, and the quant most people start with.
104,639 models have a Q4_K_S file on ggufmodels.com, and those files range from 0.0 to 598.4 GB. The most downloaded are Qwen3.8-27B, Qwen3.8-27B and Tiel Coder 35B A3B GGUF MTP. For a longer explanation of every quant type, read GGUF quantization explained.
Frequently asked questions
What is Q4_K_S in GGUF?
4-bit k-quant, small. Slightly smaller and a little lower quality than Q4_K_M. It averages about 4.6 bits per weight.
Should I use Q4_K_S?
The usual sweet spot between file size and quality, and the quant most people start with.
How big is a Q4_K_S file?
Q4_K_S files on this site range from 0.0 to 598.4 GB, depending on the model. As a rule of thumb it is about 29% of the 16-bit size.