IQ3_XXS GGUF models

A GGUF quantization format. Lower numbers mean smaller files and more quality loss. 28129 models have a IQ3_XXS file. Sizes run from 2M to 2450B parameters. The longest trained context is 10240K tokens.

Most downloadedNewestLargestLongest contextSmallest
LL
Llama 3.2 3B Instruct
Llama · DavidAU
Params3.21B
Context128K
Q4_K_M2.0 GB
Roleplay
QW
Huihui Qwen3 Coder Next Abliterated
Qwen · mradermacher
Params79.7B
Context256K
Q4_K_M48.5 GB
Coding
OL
OLMoE 1B 7B 0924 Instruct
OLMo · mradermacher
Params6.9B
Context4K
Q4_K_M4.2 GB
GGUF
QW
Qwen3.8-27B
Qwen · ahmed-amedo
Params27.3B
Context256K
Q4_K_M16.6 GB
Vision
CY
Cydonia 24B v4.3 Absolute Heresy
Cydonia · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
GGUF
QW
Qwen3-Vl-30B-A3B-Thinking
Qwen · unsloth
Params30.5B
Context256K
Q4_K_M18.6 GB
ReasoningVision
NE
nexus legal
Nexus · King3Djbl
Params1.54B
Context32K
Q4_K_M1.0 GB
GGUF
QW
Qwen2.5 VL 32B Instruct
Qwen · mradermacher
Params32.8B
Context125K
Q4_K_M19.9 GB
Vision
GR
grok 2
Grok · unsloth
Params270B
Context128K
Q4_K_M164.1 GB
GGUF
MO
Moonlight 16B A3B Instruct
Moonlight · mmnga
Params16B
Context8K
Q4_K_M10.5 GB
GGUF
GE
Gemma-3N-E2B-It
Gemma · unsloth
Params4.46B
Context32K
Q4_K_M3.0 GB
Vision
QW
Qwen2.5-VL-7B-Instruct
Qwen · Mungert
Params7.6B
Context125K
Q4_K_M4.8 GB
Vision
QW
Qwen3.8 27B Heretic Abliterated Uncensored
Qwen · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
GGUF
TI
TildeOpen 30B Instruct LV
Tildeopen · mradermacher
Params30.7B
Context64K
Q4_K_M18.5 GB
GGUF
QW
Qwen3.5 4B MultiDistill 3xMerge UNCENSORED Heretic
Qwen · mradermacher
Params4.21B
Context256K
Q4_K_M2.7 GB
Reasoning
FR
FrogNano 4B Heretic
Frognano · mradermacher
Params4.21B
Context256K
Q4_K_M2.7 GB
CodingRoleplayTool use
QW
qwen image edit
Qwen · calcuis
Params20.4B
Context—
Q4_K_M11.7 GB
GGUF
QW
Qwen3.8-27B
Qwen · aiartorious
Params27.3B
Context256K
Q4_K_M16.6 GB
Vision
QW
Qwen3.5 2B Base Heretic
Qwen · mradermacher
Params1.88B
Context256K
Q4_K_M1.3 GB
GGUF
NE
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B
Nemotron · AtomicChat
Params124B
Context1024K
Smallest45.7 GB
ReasoningTool useVision
GR
granite-8b-code-instruct
Granite · mradermacher
Params8.05B
Context4K
Q4_K_M4.9 GB
CodingTool use
YI
Yi-1.5-9B-Chat
Yi · bartowski
Params8.83B
Context4K
Q4_K_M5.3 GB
GGUF
QW
Qwen3-8B-128K
Qwen · unsloth
Params8.19B
Context128K
Q4_K_M5.0 GB
GGUF
OR
Ornith1.0 9B 3MPER0RR Abliterated
Ornith · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
ReasoningTool use
GL
GLM Z1 32B 0414
GLM · bartowski
Params32.6B
Context32K
Q4_K_M19.7 GB
GGUF
QW
Qwen3 VL Reranker 2B
Qwen · mradermacher
Params1.72B
Context256K
Q4_K_M1.1 GB
EmbeddingVision
HA
L3.1 70B Hanami X1
Hanami · bartowski
Params70.6B
Context128K
Q4_K_M42.5 GB
GGUF
MI
Mistral-Large-4-1T-A52B-DSpark
Mistral · unsloth
Params1050B
Context1024K
Smallest2.2 GB
ReasoningTool useVision
MI
Codestral 22B v0.1 IMat
Mistral · legraphista
Params22.2B
Context32K
Smallest4.8 GB
Coding
DE
DeepSeek V2.5
DeepSeek · legraphista
Params236B
Context160K
Smallest47.4 GB
GGUF

Other quants

Keep browsing