GGUF models by Edge-Quant

79 GGUF models quantized by Edge-Quant on Hugging Face. Sizes run from 16M to 9.4B parameters. The longest trained context is 501K tokens.

Most downloadedNewestLargestLongest contextSmallest
NA
Nanbeige4.1 3B
Nanbeige · Edge-Quant
Params3.93B
Context192K
Q4_K_M2.4 GB
GGUF
NA
Nanbeige4.1 3B
Nanbeige · Edge-Quant
Params3.93B
Context192K
Smallest4.2 GB
ReasoningTool use
GE
Gemmasutra Mini 2B v1
Gemma · Edge-Quant
Params2.61B
Context8K
Q4_K_M1.7 GB
GGUF
GL
GLM 4.6V Flash
GLM · Edge-Quant
Params9.4B
Context128K
Q4_K_M6.2 GB
Vision
GE
Gemma 3n E2B It
Gemma · Edge-Quant
Params4.46B
Context32K
Q4_K_M2.8 GB
Vision
QW
Dolphin3.0 Qwen2.5 1.5B
Qwen · Edge-Quant
Params1.54B
Context128K
Q4_K_M1.0 GB
CodingTool use
GE
Gemma 3n E4B It
Gemma · Edge-Quant
Params6.87B
Context32K
Q4_K_M4.2 GB
Vision
DO
Dolphin v2
Dolphin · Edge-Quant
Params3.09B
Context125K
Q4_K_M1.9 GB
CodingTool useVision
QW
Qwen3 VL 2B
Qwen · Edge-Quant
Params1.72B
Context256K
Q4_K_M1.1 GB
Tool useVision
QW
Qwen3 4B Instruct 2507
Qwen · Edge-Quant
Params4.02B
Context256K
Q4_K_M2.5 GB
Tool use
DA
H2O Danube3 500m Chat
Danube · Edge-Quant
Params514M
Context8K
Q4_K_M0.3 GB
GGUF
GL
Glm Edge 1.5b Chat
GLM · Edge-Quant
Params1.59B
Context2K
Q4_K_M1.0 GB
GGUF
SA
Sarvam 1
Sarvam · Edge-Quant
Params2.53B
Context8K
Q4_K_M1.5 GB
Tool use
MI
Ministral 3 3B Reasoning 2512
Mistral · Edge-Quant
Params3.43B
Context256K
Q4_K_M2.1 GB
Reasoning
HU
Hunyuan 7B Instruct
Hunyuan · Edge-Quant
Params7.5B
Context256K
Q4_K_M4.6 GB
GGUF
JA
Jan v1 4B
Jan · Edge-Quant
Params4.02B
Context256K
Q4_K_M2.5 GB
GGUF
NE
AceReason Nemotron 1.1 7B
Nemotron · Edge-Quant
Params7.62B
Context128K
Q4_K_M4.7 GB
CodingReasoningTool use
HY
HY MT1.5 1.8B
Edge-Quant
Params1.79B
Context256K
Q4_K_M1.1 GB
GGUF
EU
EuroLLM 1.7B Instruct Q4 K M
Eurollm · Edge-Quant
Params1.66B
Context4K
Q4_K_M1.0 GB
GGUF
OL
Olmo 3 7B Instruct
OLMo · Edge-Quant
Params7.3B
Context64K
Q4_K_M4.5 GB
GGUF
GR
Granite 3.3 2b Instruct
Granite · Edge-Quant
Params2.53B
Context128K
Q4_K_M1.5 GB
ReasoningTool use
OL
OLMo 1B 0724 Hf
OLMo · Edge-Quant
Params1.28B
Context4K
Smallest1.1 GB
GGUF
HU
Huihui MoE 1.2B A0.6B
Huihui · Edge-Quant
Params1.28B
Context40K
Q4_K_M0.8 GB
GGUF
HU
Hunyuan 1.8B Instruct
Hunyuan · Edge-Quant
Params1.79B
Context256K
Q4_K_M1.1 GB
ReasoningTool use
HU
Hunyuan 0.5B Instruct
Hunyuan · Edge-Quant
Params539M
Context256K
Smallest0.6 GB
ReasoningTool use
MI
MiniCPM4.1 8B
MiniCPM · Edge-Quant
Params8.19B
Context64K
Q4_K_M5.0 GB
Reasoning
MI
MicroThinker 1B Preview
Microthinker · Edge-Quant
Params1.24B
Context128K
Smallest0.9 GB
ReasoningTool use
LF
LFM2 1.2B
Lfm · Edge-Quant
Params1.17B
Context125K
Q4_K_M0.7 GB
GGUF
ER
ERNIE 4.5 0.3B PT
Ernie · Edge-Quant
Params361M
Context128K
Q4_K_M0.2 GB
GGUF
HU
Huihui MoE 1B A0.6B SFT
Huihui · Edge-Quant
Params1.12B
Context40K
Q4_K_M0.7 GB
CodingReasoning

About Edge-Quant on Hugging Face

Edge-Quant has published 79 GGUF models that ggufmodels.com tracks, across families such as Gemma, Qwen and Nanbeige. The most downloaded are Nanbeige4.1 3B, Nanbeige4.1 3B and Gemmasutra Mini 2B v1.

The quants Edge-Quant publishes most often are Q4_K_M (67), Q8_0 (7), Q5_K_M (3) and Q6_K (2). Their Q4_K_M files range from 0.2 to 6.2 GB. Parameter counts run from 16M to 9.4B, and the longest trained context window is 501K tokens.

Frequently asked questions

What GGUF models has Edge-Quant published?

79 models are listed here, mainly from the Gemma, Qwen and Nanbeige families. The most downloaded is Nanbeige4.1 3B.

Which quants does Edge-Quant publish?

The most common are Q4_K_M (67 models), Q8_0 (7 models), Q5_K_M (3 models) and Q6_K (2 models).

Keep browsing