GGUF models by noctrex
173 GGUF models quantized by noctrex on Hugging Face. Sizes run from 1M to 396B parameters. The longest trained context is 10240K tokens.
About noctrex on Hugging Face
noctrex has published 173 GGUF models that ggufmodels.com tracks, across families such as Qwen, Huihui and GLM. The most downloaded are Huihui Qwen3 VL 4B Instruct Abliterated, Huihui Qwen3 VL 30B A3B Instruct Abliterated and Huihui Qwen3 VL 8B Instruct Abliterated.
The quants noctrex publishes most often are MXFP4_MOE (120), Q8_0 (47), Q6_K (47) and Q5_K_M (46). Their Q4_K_M files range from 0.4 to 19.8 GB. Parameter counts run from 1M to 396B, and the longest trained context window is 10240K tokens.
Frequently asked questions
What GGUF models has noctrex published?
173 models are listed here, mainly from the Qwen, Huihui and GLM families. The most downloaded is Huihui Qwen3 VL 4B Instruct Abliterated.
Which quants does noctrex publish?
The most common are MXFP4_MOE (120 models), Q8_0 (47 models), Q6_K (47 models) and Q5_K_M (46 models).