Q6_K GGUF models
6-bit k-quant. Very close to the original model for most uses. 112449 models have a Q6_K file. Sizes run from 1M to 1030B parameters. The longest trained context is 4194304K tokens.
What is Q6_K?
6-bit k-quant. Very close to the original model for most uses. On average it stores about 6.6 bits per weight, so a Q6_K file is roughly 41% of the size of the same model in 16-bit. Very close to the original quality for most uses, at a clearly larger size than 4-bit.
112,449 models have a Q6_K file on ggufmodels.com, and those files range from 0.0 to 846.6 GB. The most downloaded are YuE2 audio tokenizer, the v9 head of Mothersuperior, on MERT-v2-FullSong, Qwen3.8-27B and Ornith-1.5-9B. For a longer explanation of every quant type, read GGUF quantization explained.
Frequently asked questions
What is Q6_K in GGUF?
6-bit k-quant. Very close to the original model for most uses. It averages about 6.6 bits per weight.
Should I use Q6_K?
Very close to the original quality for most uses, at a clearly larger size than 4-bit.
How big is a Q6_K file?
Q6_K files on this site range from 0.0 to 846.6 GB, depending on the model. As a rule of thumb it is about 41% of the 16-bit size.