GGUF models by Disya
62 GGUF models quantized by Disya on Hugging Face. Sizes run from 423M to 34.7B parameters. The longest trained context is 1000K tokens.
About Disya on Hugging Face
Disya has published 62 GGUF models that ggufmodels.com tracks, across families such as Qwen, Mistral and Hunyuan. The most downloaded are Josiefied Qwen3 30B A3B Abliterated v1, The Omega Directive Qwen3 14B v1.1 and Gemma 4 12B It Qat Q4_0 Unquantized Heretic.
The quants Disya publishes most often are Q4_K_M (24), Q8_0 (16), Q5_K_M (6) and Q4_0 (6). Their Q4_K_M files range from 5.0 to 18.6 GB. Parameter counts run from 423M to 34.7B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has Disya published?
62 models are listed here, mainly from the Qwen, Mistral and Hunyuan families. The most downloaded is Josiefied Qwen3 30B A3B Abliterated v1.
Which quants does Disya publish?
The most common are Q4_K_M (24 models), Q8_0 (16 models), Q5_K_M (6 models) and Q4_0 (6 models).