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
MI
MiniCPM5 2B Abliterated Uncensored Safetensors
MiniCPM · mradermacher
Params2.52B
Context128K
Q4_K_M1.6 GB
Tool use
QW
Cogito v1 Preview Qwen 32B
Qwen · bartowski
Params32.8B
Context128K
Q4_K_M19.8 GB
GGUF
QW
Qwen3.5 21B Claude 4.6 Opus Deckard Heretic Uncensored Thinking
Qwen · mradermacher
Params20.8B
Context256K
Q4_K_M12.9 GB
CodingReasoningRoleplay
RA
Rain-2x7B-MoE-32k-v0.2
Rain · mradermacher
Params12.9B
Context32K
Q4_K_M7.8 GB
GGUF
LL
Llama 3 Youko 70b
Llama · mradermacher
Params70.6B
Context8K
Q4_K_M42.5 GB
GGUF
MI
Mistral 2x24B MOE Power Devstral Magistral Reasoning Ultimate 44B
Mistral · mradermacher
Params43.7B
Context128K
Q4_K_M26.5 GB
CodingReasoning
AV
Avicenna 8B Base
Avicenna · mradermacher
Params8B
Context128K
Q4_K_M4.9 GB
Reasoning
QW
OPSD Qwen3 4B
Qwen · mradermacher
Params4.41B
Context40K
Q4_K_M2.7 GB
Reasoning
LL
Llama 3.1 Storm 8B
Llama · legraphista
Params8B
Context128K
Smallest2.0 GB
Reasoning
LU
Lumimaid Magnum 12B i1
Lumimaid · mradermacher
Params12.2B
Context1000K
Q4_K_M7.5 GB
GGUF
HA
Hathor_RP-v.01-L3-8B
Hathor · mradermacher
Params8B
Context8K
Q4_K_M4.9 GB
Roleplay
CH
CHIMERA 4B SFT
Chimera · mradermacher
Params4B
Context256K
Q4_K_M2.5 GB
Reasoning
MA
Marco Nano Instruct
Marco · mradermacher
Params8B
Context32K
Q4_K_M5.4 GB
Reasoning
RA
RAFT 7B
Raft · mradermacher
Params7.62B
Context128K
Q4_K_M4.7 GB
CodingReasoningTool use
AD
MN Adversary 12B v1
Adversary · mradermacher
Params12.2B
Context128K
Q4_K_M7.5 GB
Roleplay
QW
Qwable 9B Claude Fable 5
Qwable · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
GGUF
HE
HelpingAI2.5 5B
Helpingai · mradermacher
Params5.13B
Context128K
Q4_K_M3.2 GB
GGUF
PH
Phi 4 Heretic
Phi · mradermacher
Params14.7B
Context16K
Q4_K_M9.1 GB
Coding
AR
ArcaneEntanglement model64 70b i1
Arcaneentanglement · mradermacher
Params69B
Context4K
Q4_K_M41.4 GB
GGUF
LL
dolphin-2.9-llama3-70b
Llama · mradermacher
Params70.6B
Context8K
Q4_K_M42.5 GB
CodingTool use
CA
CarbonBeagle 11B i1
Carbonbeagle · mradermacher
Params10.7B
Context32K
Q4_K_M6.5 GB
GGUF
QW
Qwen3 Hermes8B v1
Qwen · mradermacher
Params8.2B
Context40K
Q4_K_M5.0 GB
ReasoningTool use
QW
Qwen3.5 27B
Qwen · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
GGUF
LL
Llama EveningMirai 3.3 70B
Llama · mradermacher
Params70.6B
Context128K
Q4_K_M42.5 GB
GGUF
SW
Swish 26B A4B
Swish · mradermacher
Params25.2B
Context256K
Q4_K_M16.8 GB
Roleplay
OL
open r1 OlympicCoder 32B
Olympiccoder · bartowski
Params32.8B
Context32K
Q4_K_M19.9 GB
CodingReasoning
OP
Openchat_3.5
Openchat · mradermacher
Params7.24B
Context8K
Q4_K_M4.4 GB
Coding
CY
Cydonia ReduX 22B v1
Cydonia · bartowski
Params22.2B
Context32K
Q4_K_M13.3 GB
GGUF
NE
Heretic Nemotron Content Safety Reasoning 4B
Nemotron · mradermacher
Params3.88B
Context128K
Q4_K_M2.5 GB
Reasoning
LL
calme-2.2-llama3.1-70b
Llama · mradermacher
Params70.6B
Context128K
Q4_K_M42.5 GB
GGUF

Other quants

Keep browsing