GGUF models by DevQuasar-13

36 GGUF models quantized by DevQuasar-13 on Hugging Face. Sizes run from 494M to 405B parameters. The longest trained context is 2096K tokens.

Most downloadedNewestLargestLongest contextSmallest
HE
NousResearch.Hermes 4 405B
Hermes · DevQuasar-13
Params405B
Context128K
Q4_K_M243.1 GB
Tool use
LL
nvidia.Llama 3 1 Nemotron Ultra 253B v1
Llama · DevQuasar-13
Params253B
Context128K
Q4_K_M150.9 GB
GGUF
DA
DavidAU.L3.1 Dark Reasoning LewdPlay Evo Hermes R1 Uncensored 8B
Dark · DevQuasar-13
Params8B
Context128K
Q4_K_M4.9 GB
ReasoningTool use
LL
Deepcogito.Cogito v1 Preview Llama 3B
Llama · DevQuasar-13
Params3.61B
Context128K
Q4_K_M2.2 GB
ReasoningTool use
QW
Huihui ai.Qwen2.5 3B Instruct Abliterated SFT
Qwen · DevQuasar-13
Params3.09B
Context32K
Q4_K_M1.9 GB
Tool use
QW
Qwen.Qwen2.5 1.5B
Qwen · DevQuasar-13
Params1.54B
Context128K
Q4_K_M1.0 GB
Tool use
GL
THUDM/GLM-Z1-32B-0414
GLM · DevQuasar-13
Params32.6B
Context32K
Q4_K_M19.7 GB
GGUF
QW
Qwen.Qwen2.5 3B Instruct
Qwen · DevQuasar-13
Params3.09B
Context32K
Q4_K_M1.9 GB
Tool use
QW
Qwen2.5 0.5B
Qwen · DevQuasar-13
Params494M
Context32K
Q4_K_M0.4 GB
Tool use
DA
DavidAU.L3.1 Dark Reasoning Unholy Hermes R1 Uncensored 8B
Dark · DevQuasar-13
Params8.03B
Context128K
Q4_K_M4.9 GB
ReasoningTool use
ZY
Zyphra.ZR1 1.5B
DevQuasar-13
Params1.78B
Context128K
Q4_K_M1.1 GB
Tool use
GL
Huihui ai.GLM 4 32B 0414 Abliterated
GLM · DevQuasar-13
Params32.6B
Context32K
Q4_K_M19.7 GB
Tool use
QW
Huihui ai.Qwen2.5 7B Instruct Abliterated SFT
Qwen · DevQuasar-13
Params7.62B
Context32K
Q4_K_M4.7 GB
Tool use
QW
Qwen.Qwen2.5 3B
Qwen · DevQuasar-13
Params3.09B
Context32K
Q4_K_M1.9 GB
Tool use
SQ
defog.sqlcoder 7b 2
Sqlcoder · DevQuasar-13
Params6.74B
Context16K
Q4_K_M4.1 GB
Coding
NE
OpenMath Nemotron 32B
Nemotron · DevQuasar-13
Params32.8B
Context128K
Q4_K_M19.9 GB
GGUF
LL
nvidia/Llama-3.1-8B-UltraLong-2M-Instruct
Llama · DevQuasar-13
Params8.04B
Context2096K
Q4_K_M4.9 GB
Tool use
NE
OpenMath Nemotron 14B
Nemotron · DevQuasar-13
Params14.8B
Context128K
Q4_K_M9.0 GB
GGUF
QW
Huihui ai.Qwen2.5 14B Instruct Abliterated SFT
Qwen · DevQuasar-13
Params14.8B
Context32K
Q4_K_M9.0 GB
Tool use
GL
THUDM/GLM-Z1-Rumination-32B-0414
GLM · DevQuasar-13
Params33B
Context128K
Q4_K_M20.0 GB
GGUF
QW
Qwen.Qwen2.5 14B Instruct
Qwen · DevQuasar-13
Params14.8B
Context32K
Q4_K_M9.0 GB
Tool use
BL
Bigscience.Bloom
Bloom · DevQuasar-13
Params180B
Context14K
Q4_K_M114.7 GB
GGUF
GL
THUDM.GLM 4 32B 0414
GLM · DevQuasar-13
Params32.6B
Context32K
Smallest12.3 GB
Tool use
NE
nvidia.AceMath RL Nemotron 7B
Nemotron · DevQuasar-13
Params7.62B
Context128K
Q4_K_M4.7 GB
ReasoningTool use
XL
xLAM 2 1b Fc R
Xlam · DevQuasar-13
Params1.54B
Context32K
Q4_K_M1.0 GB
Tool use
NE
OpenMath Nemotron 14B Kaggle
Nemotron · DevQuasar-13
Params14.8B
Context128K
Q4_K_M9.0 GB
GGUF
AS
prithivMLmods.Asterope 21 OpenR1
Asterope · DevQuasar-13
Params1.78B
Context128K
Q4_K_M1.1 GB
Tool use
NE
OpenMath Nemotron 1.5B
Nemotron · DevQuasar-13
Params1.54B
Context128K
Q4_K_M1.0 GB
GGUF
LL
defog.llama 3 sqlcoder 8b
Llama · DevQuasar-13
Params8.03B
Context8K
Q4_K_M4.9 GB
Coding
XL
xLAM 2 3b Fc R
Xlam · DevQuasar-13
Params3.09B
Context32K
Q4_K_M1.9 GB
Tool use

About DevQuasar-13 on Hugging Face

DevQuasar-13 has published 36 GGUF models that ggufmodels.com tracks, across families such as Qwen, Nemotron and Llama. The most downloaded are NousResearch.Hermes 4 405B, nvidia.Llama 3 1 Nemotron Ultra 253B v1 and DavidAU.L3.1 Dark Reasoning LewdPlay Evo Hermes R1 Uncensored 8B.

The quants DevQuasar-13 publishes most often are Q3_K_M (38), Q2_K (38), Q8_0 (37) and Q5_K_M (37). Their Q4_K_M files range from 0.4 to 243.1 GB. Parameter counts run from 494M to 405B, and the longest trained context window is 2096K tokens.

Frequently asked questions

What GGUF models has DevQuasar-13 published?

36 models are listed here, mainly from the Qwen, Nemotron and Llama families. The most downloaded is NousResearch.Hermes 4 405B.

Which quants does DevQuasar-13 publish?

The most common are Q3_K_M (38 models), Q2_K (38 models), Q8_0 (37 models) and Q5_K_M (37 models).

Keep browsing