Mergekit GGUF models
221 GGUF models in the Mergekit family, quantized by 28 uploaders. Sizes run from 323M to 34.4B parameters. The longest trained context is 1000K tokens.
About the Mergekit GGUF models
Mergekit is represented on ggufmodels.com by 221 GGUF models, quantized by 28 uploaders on Hugging Face. The most downloaded are mergekit dare ties ymiqjtz, Mergekit Dare_Ties Twfgema and mergekit slerp xlblwaw i1. Parameter counts run from 323M to 34.4B, and the longest trained context window is 1000K tokens.
A Q4_K_M file of a Mergekit model is between 0.5 to 41.3 GB, which is the quant most people start with. Most files come from RichardErkhov (66), mradermacher (38) and fengpeisheng1 (15). The quants you will meet most often are Q4_K_M (129), Q2_K (126), Q3_K_M (118) and Q5_K_M (109), where the number is how many models offer that quant.
Frequently asked questions
Which Mergekit GGUF should I download?
The most downloaded Mergekit GGUF here is mergekit dare ties ymiqjtz. Pick the size that fits your GPU first, then choose a quant: Q4_K_M is the usual starting point, and Q5_K_M or Q6_K give more quality if you have memory to spare. The VRAM pages below list what fits each GPU.
How much VRAM does a Mergekit model need?
Q4_K_M files of Mergekit models on this site range from 0.5 to 41.3 GB. Add the KV cache for your context length and about 0.5 GB for the runtime, which usually means 1.5 to 3 GB on top of the file at an 8K context.
Where do Mergekit GGUF files come from?
The most common quantizers here are RichardErkhov, mradermacher and fengpeisheng1. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.