GGUF models by Theros
103 GGUF models quantized by Theros on Hugging Face. Sizes run from 40M to 14.8B parameters. The longest trained context is 1000K tokens.
QW
Qwen2.5 ColdBrew Aetheria Test3
Qwen · Theros
Params7.61B
Context128K
Smallest5.4 GB
CO
ColdBrew Nemo 12B Arcane Base M
Coldbrew · Theros
Params12.2B
Context128K
Q4_K_M7.5 GB
CO
L3 ColdBrew 250 Q4 K M
Coldbrew · Theros
Params8.03B
Context8K
Q4_K_M4.9 GB
CO
Q2.5 ColdBrew RxV
Coldbrew · Theros
Params9.24B
Context128K
Q4_K_M5.7 GB
WA
WayChron 12B Test1
Waychron · Theros
Params12.2B
Context1000K
Q4_K_M7.5 GB
CO
L3 ColdBrew R1 test1 Q4 K M
Coldbrew · Theros
Params8.03B
Context8K
Q4_K_M4.9 GB
CO
ColdBrew 600 4bit F16
Coldbrew · Theros
Params40M
Context—
Smallest0.2 GB
CO
ColdBrew Nemo 12B Arcane Base Test1
Coldbrew · Theros
Params12.2B
Context1000K
Q4_K_M7.5 GB
SO
SOR ColdBrew 12B Base Test2
Sor · Theros
Params12.2B
Context1000K
Q4_K_M7.5 GB
GE
Gemma3 ColdBrew Lorenz 12b Test1
Gemma · Theros
Params11.8B
Context128K
Smallest8.4 GB
CO
Q2.5 ColdBrew RxVm Test0
Coldbrew · Theros
Params9.24B
Context128K
Smallest6.6 GB
GE
Gemma3 ColdBrew Lorenz
Gemma · Theros
Params11.8B
Context128K
Smallest8.4 GB
QW
Qwen3 CB14B
Qwen · Theros
Params14.8B
Context40K
Q4_K_M9.0 GB