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
MI
MiniAI Quata2 4b
Miniai · mradermacher
Params4B
Context40K
Q4_K_M2.5 GB
Reasoning
DE
DeepSeek R1 Distill Qwen 7B
DeepSeek · mradermacher
Params7.62B
Context128K
Q4_K_M4.7 GB
Reasoning
GL
Q3.6 27B GLM 5.1 DA
GLM · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
Reasoning
UP
upstage SOLAR 10.7B Instruct v1.0
Upstage · RichardErkhov
Params10.7B
Context4K
Q4_K_M6.5 GB
GGUF
QW
JoyFox Qwen3.6 35B A3B RP Aggressive
Qwen · mradermacher
Params35.5B
Context256K
Q4_K_M21.7 GB
Roleplay
AR
ArmenianGPT 1.0 3B
Armeniangpt · mradermacher
Params3.43B
Context256K
Q4_K_M2.1 GB
Reasoning
ST
L3-8B-Stheno-v3.3-32K
Stheno · Lewdiculous
Params8B
Context32K
Q4_K_M4.9 GB
Roleplay
LL
Llama3-ChatQA-1.5-70B
Llama · mradermacher
Params70.6B
Context8K
Q4_K_M42.5 GB
GGUF
FI
Fireplace 34b
Fireplace · mradermacher
Params34.4B
Context195K
Q4_K_M22.1 GB
CodingTool use
LL
Llama 3.3 70B Instruct ablated i1
Llama · mradermacher
Params70.6B
Context128K
Q4_K_M42.5 GB
GGUF
GR
Granite 8b Code Instruct 4k
Granite · mradermacher
Params8.05B
Context4K
Q4_K_M4.9 GB
CodingTool use
QW
Qwen2.5 Wolverine CODER 11B V2
Qwen · mradermacher
Params10.9B
Context32K
Q4_K_M6.6 GB
Coding
LL
Openthai Llama Pretrained 7B
Llama · RichardErkhov
Params6.94B
Context4K
Q4_K_M4.2 GB
GGUF
GE
Gemma4 Gutenberg Literotica 12B
Gemma · mradermacher
Params11.9B
Context256K
Q4_K_M7.4 GB
GGUF
GE
Gemma 2 27b It
Gemma · ThomasBaruzier
Params27.2B
Context8K
Q4_K_M16.6 GB
GGUF
ME
L3.1 70b MeowMix i1
Meowmix · mradermacher
Params70.6B
Context128K
Q4_K_M42.5 GB
GGUF
CO
C4AI orpo model
Command R · mradermacher
Params135M
Context2K
Q4_K_M0.1 GB
GGUF
ST
Storm-7B
Storm · mradermacher
Params7.24B
Context8K
Q4_K_M4.4 GB
GGUF
PH
Phind 70B
Phind · mradermacher
Params70.6B
Context128K
Q4_K_M42.5 GB
Coding
YI
Yi-34B-200K-AEZAKMI-v2
Yi · mradermacher
Params34.4B
Context195K
Q4_K_M20.7 GB
GGUF
LL
llama-3-8b-samantha
Llama · mradermacher
Params8B
Context8K
Q4_K_M4.9 GB
GGUF
GR
Granite 20b Code Instruct 8k
Granite · mradermacher
Params20B
Context8K
Q4_K_M12.8 GB
CodingTool use
SO
SOLAR 10.7B Instruct v1.0
Solar · mradermacher
Params10.7B
Context4K
Q4_K_M6.5 GB
GGUF
HA
HarmonicHarlequin v5 20B
Harmonicharlequin · mradermacher
Params33.3B
Context4K
Q4_K_M20.1 GB
GGUF
MI
Infinity Instruct 7M Gen Mistral 7B
Mistral · mradermacher
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
MI
Mixtral 8x22B v0.1
Mistral · LiteLLMs
Params141B
Context64K
Q4_K_M85.6 GB
GGUF
DE
DeepSeek-V2-Lite
DeepSeek · mradermacher
Params15.7B
Context160K
Q4_K_M10.4 GB
GGUF
LL
Code-Llama-2-13B-instruct-text2sql
Llama · mradermacher
Params13B
Context16K
Q4_K_M7.9 GB
Coding
PS
Psyonic-Rose-20B-Higher-Quality
Psyonic · mradermacher
Params20B
Context4K
Q4_K_M12.0 GB
GGUF
CO
CodeQwen1.5-7B
Codeqwen · bartowski
Params7.25B
Context64K
Q4_K_M4.7 GB
GGUF

Other quants

Keep browsing