IQ4_XS GGUF models

Importance-matrix 4-bit quant, extra small. Small files with good quality for the size. 97368 models have a IQ4_XS file. Sizes run from 1M to 2450B parameters. The longest trained context is 102400K tokens.

Most downloadedNewestLargestLongest contextSmallest
NE
NeoHorse-1-4B
Neohorse · ngquocvinh
Params4.21B
Context256K
Q4_K_M2.7 GB
ReasoningTool use
QW
Qwen3.8 27B Pi
Qwen · bartowski
Params27.3B
Context256K
Q4_K_M17.4 GB
CodingReasoningTool use
LF
Lfm2.5-8B-A1B
Lfm · unsloth
Params8.47B
Context125K
Q4_K_M5.3 GB
GGUF
LF
Lfm2.5-Vl-3B
Lfm · unsloth
Params2.7B
Context125K
Q4_K_M1.7 GB
Vision
QW
Qwen3 VL 8B GLM 4.7 Flash Heretic Uncensored Thinking
Qwen · mradermacher
Params8.19B
Context256K
Q4_K_M5.0 GB
ReasoningRoleplayVision
QW
Qwen2.5 Math 7B Instruct
Qwen · bartowski
Params7.62B
Context4K
Q4_K_M4.7 GB
GGUF
MI
MiMo V2.6 Flash MOPD
Mimo · spiritfather
Params310B
Context1024K
Q4_K_M189.5 GB
GGUF
MI
Ministral 3 3B Instruct 2512
Mistral · bartowski
Params3.43B
Context256K
Q4_K_M2.1 GB
Vision
QW
Qwen3.5 122B A10B
Qwen · bartowski
Params125B
Context256K
Q4_K_M77.6 GB
Vision
WI
Wicked Tutu 31B v0.1
Wicked · ReadyArt
Params30.7B
Context256K
Q4_K_M18.7 GB
Roleplay
IM
z image
Image · gguf-org
Params84M
Context—
Q4_K_M4.5 GB
GGUF
CL
Clef Flash Heretic
Clef · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
Vision
LL
Llama Joycaption Beta One Hf Llava
Llama · mradermacher
Params8B
Context128K
Q4_K_M4.9 GB
Vision
MI
MiniCPM-V-4.7-35B-A3B-Uncensored
MiniCPM · gbuzhf
Params34.7B
Context256K
Smallest11.9 GB
Vision
QW
Qwen3.5-397B-A17B
Qwen · unsloth
Params397B
Context256K
Q4_K_M244.1 GB
Vision
QW
Qwen3.8 27B ZeroRefusal UD IQ4 XS MTP
Qwen · QQZ2026
Params461M
Context—
Smallest0.0 GB
GGUF
LL
dolphin 2.9 llama3 8b
Llama · bartowski
Params8B
Context8K
Q4_K_M4.9 GB
CodingTool use
QW
Qwen3 0.6B Abliterated
Qwen · mradermacher
Params596M
Context40K
Q4_K_M0.4 GB
GGUF
QW
Qwen3 0.6B Heretic Abliterated Uncensored
Qwen · mradermacher
Params596M
Context40K
Q4_K_M0.4 GB
GGUF
MI
Ministral-3-8B-Instruct-2512
Mistral · unsloth
Params8.49B
Context256K
Q4_K_M5.2 GB
Vision
QW
Qwen3 0.6B Heretic
Qwen · mradermacher
Params596M
Context40K
Q4_K_M0.4 GB
ReasoningTool use
QW
Qwen3.6 35B A3B Uncensored
Qwen · xCloudinfo
Params34.7B
Context256K
Q4_K_M21.2 GB
ReasoningTool use
LL
Hermes 3 Llama 3.2 3b Base Fft Chatml Rerun 8xh100 E4 2024 11 24 4k Steps
Llama · bartowski
Params3.21B
Context128K
Q4_K_M2.0 GB
RoleplayTool use
QW
Bahushruth Qwen3.6 35B A3B Abliterated v4
Qwen · Bahushruth
Params34.7B
Context256K
Q4_K_M21.2 GB
GGUF
QW
Qwen3 14B Abliterated Conservative
Qwen · richardyoung
Params14.8B
Context40K
Q4_K_M9.0 GB
GGUF
QW
Qwen3 Next 80B A3B Instruct
Qwen · bartowski
Params80B
Context256K
Q4_K_M48.7 GB
GGUF
NE
NVIDIA-Nemotron-3.5-Bolt-74B-A9B
Nemotron · AtomicChat
Params74B
Context256K
Q4_K_M51.7 GB
GGUF
QW
Qwen3.5-2B
Qwen · unsloth
Params1.94B
Context256K
Q4_K_M1.3 GB
Vision
QW
Z Image Qwen 3 4b text encoder
Qwen · worstplayer
Params4B
Context40K
Smallest2.1 GB
Coding
LL
testmod koen llama2 13b dpotrain testver
Llama · RichardErkhov
Params13B
Context—
Q4_K_M0.0 GB
GGUF

Other quants

Keep browsing