Moss GGUF models
85 GGUF models in the Moss family, quantized by 52 uploaders. Sizes run from 7M to 11.3B parameters. The longest trained context is 256K tokens.
About the Moss GGUF models
Moss is represented on ggufmodels.com by 85 GGUF models, quantized by 52 uploaders on Hugging Face. The most downloaded are MOSS Music 8B Instruct, MOSS-TTS-v1.5 and MOSS-TTSD-v1.0. Parameter counts run from 7M to 11.3B, and the longest trained context window is 256K tokens.
A Q4_K_M file of a Moss model is between 0.0 to 5.8 GB, which is the quant most people start with. Most files come from vokra (9), axiomofmind (7) and cstr (5). The quants you will meet most often are Q8_0 (39), F16 (30), Q4_K_M (23) and Q6_K (20), where the number is how many models offer that quant.
Frequently asked questions
Which Moss GGUF should I download?
The most downloaded Moss GGUF here is MOSS Music 8B Instruct. 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 Moss model need?
Q4_K_M files of Moss models on this site range from 0.0 to 5.8 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 Moss GGUF files come from?
The most common quantizers here are vokra, axiomofmind and cstr. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.