IQ3_S GGUF models

A GGUF quantization format. Lower numbers mean smaller files and more quality loss. 45340 models have a IQ3_S file. Sizes run from 1M to 1030B parameters. The longest trained context is 102400K tokens.

Most downloadedNewestLargestLongest contextSmallest
LL
Llama3.2 30B A3B II Dark Champion INSTRUCT Heretic Abliterated Uncensored
Llama · mradermacher
Params30B
Context128K
Q4_K_M18.4 GB
Roleplay
PH
Tess v2.5 Phi 3 medium 128k 14B i1
Phi · mradermacher
Params14B
Context128K
Q4_K_M8.6 GB
GGUF
GL
GLM 4.5V
GLM · mradermacher
Params107B
Context64K
Q4_K_M70.4 GB
GGUF
MI
MiniCPM5 2B Claude Fable5 heretic
MiniCPM · saidutta69
Params2.52B
Context128K
Q4_K_M1.6 GB
ReasoningTool use
MI
Mistral dolphin mix cine open Ne NSFW
Mistral · mradermacher
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
EU
Eurus 70b sft fixed i1
Eurus · mradermacher
Params69B
Context16K
Q4_K_M41.4 GB
Reasoning
SW
Swallow-7b-hf
Swallow · mradermacher
Params6.83B
Context4K
Q4_K_M4.1 GB
GGUF
MI
Hermes 2 Pro Mistral 7B
Mistral · backyardai
Params7.24B
Context32K
Q4_K_M4.4 GB
Tool use
QW
Qwen3.5 0.8B Japanese SFT v2
Qwen · mradermacher
Params752M
Context256K
Q4_K_M0.5 GB
GGUF
RY
RynnUnify Agentic Reasoner
Rynnunify · mradermacher
Params8.19B
Context256K
Q4_K_M5.0 GB
Tool use
MU
Muse Glimmer 30B pruned V2.5 en pl
Muse · bluevoid-pl
Params27B
Context128K
Q4_K_M16.3 GB
ReasoningTool useVision
MI
chronos mistral 7b i1
Mistral · mradermacher
Params7.24B
Context8K
Q4_K_M4.4 GB
GGUF
DE
DeepSeek R1 Distill Qwen 14B 3MPER0RR Abliterated
DeepSeek · mradermacher
Params14.8B
Context128K
Q4_K_M9.0 GB
Reasoning
QW
Qwen3 8B ERP v0.1
Qwen · Aratako
Params8.19B
Context40K
Q4_K_M5.0 GB
GGUF
NE
NeuralMona MoE 4x7B i1
Neuralmona · mradermacher
Params24.2B
Context32K
Q4_K_M14.8 GB
GGUF
KA
KAI 7B Instruct v0.1
Kai · mradermacher
Params7.24B
Context32K
Q4_K_M4.4 GB
Coding
PS
Psyonic-Cetacean-full-precision
Psyonic · mradermacher
Params20B
Context4K
Q4_K_M12.0 GB
GGUF
VE
Veritas 8B Fact Checker Non Thinking 1.0
Veritas · mradermacher
Params8.19B
Context40K
Q4_K_M5.0 GB
Reasoning
QW
Qwen2.5 Coder 7B Instruct
Qwen · CISCai
Params7.62B
Context32K
Smallest1.9 GB
Coding
HU
HumanlikeRP
Humanlikerp · mradermacher
Params8.83B
Context12K
Q4_K_M5.3 GB
GGUF
QW
Qwen3.5 4B EmperoAI Qwen3.8 Distill Heretic Abliterated
Qwen · mradermacher
Params4.21B
Context256K
Q4_K_M2.7 GB
ReasoningTool use
LL
LlamaScientific 8B
Llama · mradermacher
Params8.03B
Context128K
Q4_K_M4.9 GB
Tool use
JI
Jina Reranker v3.5
Jina · jinaai
Params596M
Context128K
Q4_K_M0.4 GB
Embedding
QW
Agrillm Qwen3 30B A3B
Qwen · mradermacher
Params30.5B
Context256K
Q4_K_M18.6 GB
GGUF
QW
Co Rewarding I Qwen3 8B Base MATH
Qwen · mradermacher
Params8.19B
Context32K
Q4_K_M5.0 GB
GGUF
FA
Falcon H1 34B Instruct
Falcon · tiiuae
Params33.6B
Context256K
Q4_K_M20.3 GB
GGUF
SO
Socio Foundation 8B
Socio · mradermacher
Params8.19B
Context40K
Q4_K_M5.0 GB
GGUF
BA
bagel dpo 20b v04 i1
Bagel · mradermacher
Params19.9B
Context32K
Q4_K_M12.7 GB
CodingRoleplayTool use
QW
Qwen3 Coder 30B A3B Instruct Westernized
Qwen · mradermacher
Params30.5B
Context256K
Q4_K_M18.6 GB
Coding
IN
InnerIAI chat 7b grok
Inneriai · mradermacher
Params7.24B
Context32K
Q4_K_M4.5 GB
GGUF

Other quants

Keep browsing