IQ2_XS GGUF models

A GGUF quantization format. Lower numbers mean smaller files and more quality loss. 26336 models have a IQ2_XS file. Sizes run from 2M to 2450B parameters. The longest trained context is 10240K tokens.

Most downloadedNewestLargestLongest contextSmallest
QW
Qwen3.8 Flash Next GSQ RCO
Qwen · ISTA-DASLab
Params177B
Context256K
Smallest66.4 GB
Vision
QW
Qwen3.8 Flash Next GSQ RCO abliterated
Qwen · SC117
Params177B
Context256K
Smallest0.9 GB
GGUF
QW
Qwen3.8-27B
Qwen · huihui-ai
Params27.3B
Context256K
Q4_K_M16.6 GB
Vision
DE
DeepSeek V4.1 Flash 1017 GSQ RCO
DeepSeek · ISTA-DASLab
Params197B
Context—
Smallest110.6 GB
GGUF
QW
Swift 1.5 Qwen3.8 27B GSQ RCO
Qwen · ukisai
Params27B
Context—
Smallest0.0 GB
Reasoning
QW
Qwen3.8-27B
Qwen · peculiar-ragdoll
Params27.3B
Context256K
Q4_K_M16.5 GB
Vision
QW
Qwen3.8 27B GSQ RCO
Qwen · ISTA-DASLab
Params27B
Context—
Smallest0.0 GB
Vision
QW
Qwen2.5 32B Instruct
Qwen · bartowski
Params32.8B
Context32K
Q4_K_M19.9 GB
GGUF
CY
Cyber Ornith 1.5 9B OBLITERATED
Cyber · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
CodingReasoningTool use
MU
Muse Glimmer 30B
Muse · unsloth
Params27.9B
Context128K
Smallest1.6 GB
Vision
QW
Swift 1.5 Qwen3.8 Flash Next GSQ RCO
Qwen · ukisai
Params177B
Context—
Smallest0.6 GB
ReasoningVision
QW
Qwen3.8 27B
Qwen · DavidAU
Params27B
Context256K
Q4_K_M16.5 GB
CodingReasoningVision
QW
Swift Qwen3.8 27b
Qwen · ukisai
Params27.3B
Context256K
Q4_K_M18.0 GB
ReasoningVision
YU
YuE2 LM
Yue · scragnog
Params3.8B
Context24K
Q4_K_M2.4 GB
GGUF
QW
Swift 1.5 Qwen3.8 Flash Next GSQ RCO abliterated
Qwen · SC117
Params177B
Context256K
Smallest0.9 GB
GGUF
BI
BigBang v1
Bigbang · bartowski
Params35.5B
Context256K
Q4_K_M21.9 GB
Vision
QW
Huihui Qwen3.8 Flash Next abliterated
Qwen · huihui-ai
Params177B
Context256K
Smallest68.2 GB
ReasoningVision
TH
Thomson 1.0 Small
Thomson · bartowski
Params34.7B
Context256K
Q4_K_M21.4 GB
Vision
QW
Qwen3.8 Flash Next abliterated GSQ RCO Strata
Qwen · alesha-pro
Params—
Context—
Smallest0.0 GB
GGUF
KA
KAT Coder V2.5 Dev
Kat · bartowski
Params34.7B
Context256K
Q4_K_M21.4 GB
Coding
QW
Qwen3.8 27B
Qwen · bartowski
Params27.3B
Context256K
Q4_K_M17.4 GB
Vision
QW
Huihui Qwen3.8 Flash Next abliterated v2
Qwen · huihui-ai
Params177B
Context—
Smallest1.2 GB
ReasoningVision
LI
Ling 3.0 Tiny
Ling · bloomer010
Params7.89B
Context128K
Q4_K_M4.8 GB
Coding
LI
Ling 3.0 Flash
Ling · AtomicChat
Params124B
Context128K
Q4_K_M75.3 GB
ReasoningTool use
QW
Swift 1.5 Qwen3.8 27B
Qwen · ukisai
Params27.3B
Context256K
Q4_K_M17.4 GB
ReasoningVision
GE
humanizer-12b
Gemma · jialinyyzz
Params11.9B
Context256K
Q4_K_M7.6 GB
Vision
QW
Qwen3.6 35B A3B
Qwen · bartowski
Params35.5B
Context256K
Q4_K_M22.3 GB
Vision
QW
Swift 1.5 Qwen3.8 Flash Next
Qwen · ukisai
Params177B
Context256K
Q4_K_M119.6 GB
ReasoningVision
QW
Qwen3.8 27B Uncensored
Qwen · bartowski
Params27.3B
Context256K
Q4_K_M17.8 GB
ReasoningTool useVision
TA
UI-TARS-1.5-7B
Tars · Mungert
Params7.6B
Context125K
Q4_K_M4.8 GB
Vision

What is IQ2_XS?

A GGUF quantization format. Lower numbers mean smaller files and more quality loss. On average it stores about 2.3 bits per weight, so a IQ2_XS file is roughly 14% of the size of the same model in 16-bit. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

26,336 models have a IQ2_XS file on ggufmodels.com, and those files range from 0.0 to 730.7 GB. The most downloaded are Qwen3.8 Flash Next GSQ RCO, Qwen3.8 Flash Next GSQ RCO abliterated and Qwen3.8-27B. For a longer explanation of every quant type, read GGUF quantization explained.

Frequently asked questions

What is IQ2_XS in GGUF?

A GGUF quantization format. Lower numbers mean smaller files and more quality loss. It averages about 2.3 bits per weight.

Should I use IQ2_XS?

Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

How big is a IQ2_XS file?

IQ2_XS files on this site range from 0.0 to 730.7 GB, depending on the model. As a rule of thumb it is about 14% of the 16-bit size.

Other quants

Keep browsing