Q5_K_M GGUF models
5-bit k-quant, medium. A step up in quality from Q4_K_M for a modest size increase. 117156 models have a Q5_K_M file. Sizes run from 1M to 1570B parameters. The longest trained context is 4194304K tokens.
What is Q5_K_M?
5-bit k-quant, medium. A step up in quality from Q4_K_M for a modest size increase. On average it stores about 5.7 bits per weight, so a Q5_K_M file is roughly 36% of the size of the same model in 16-bit. Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.
117,156 models have a Q5_K_M file on ggufmodels.com, and those files range from 0.0 to 1115.6 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 Q5_K_M in GGUF?
5-bit k-quant, medium. A step up in quality from Q4_K_M for a modest size increase. It averages about 5.7 bits per weight.
Should I use Q5_K_M?
Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.
How big is a Q5_K_M file?
Q5_K_M files on this site range from 0.0 to 1115.6 GB, depending on the model. As a rule of thumb it is about 36% of the 16-bit size.