IQ2_XXS GGUF models

Importance-matrix 2-bit quant. Extreme compression with a large quality cost. 26647 models have a IQ2_XXS file. Sizes run from 2M to 2780B parameters. The longest trained context is 10240K tokens.

Most downloadedNewestLargestLongest contextSmallest
QW
Qwen3.5 Queen 27B
Qwen · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
RoleplayVision
GE
Gemma-3-4B-It
Gemma · zw89
Params3.88B
Context128K
Q4_K_M2.5 GB
Vision
IS
Ising Calibration 1 35B A3B
Ising · mradermacher
Params34.7B
Context256K
Q4_K_M21.2 GB
Vision
SH
Shieldgemma 27b
Shieldgemma · mradermacher
Params27.2B
Context8K
Q4_K_M16.6 GB
GGUF
LL
Llama Guard 3 8B
Llama · mradermacher
Params8B
Context128K
Q4_K_M4.9 GB
GGUF
WO
WoonaV1.2 9b i1
Woonav · mradermacher
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
FA
Falcon3 7B Instruct Heretic
Falcon · mradermacher
Params7.46B
Context32K
Q4_K_M4.6 GB
GGUF
MO
MoMo 70B V1.1 i1
Momo · mradermacher
Params69B
Context4K
Q4_K_M41.4 GB
GGUF
QW
Qwen3.5 27B Marvin DPO V2
Qwen · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
Roleplay
QW
Qwen2 VL 72B Instruct
Qwen · bartowski
Params72.7B
Context32K
Q4_K_M47.4 GB
Vision
LL
Meta Llama 3 8B Instruct
Llama · mradermacher
Params8B
Context8K
Q4_K_M4.9 GB
Tool use
QW
bu-30b-a3b-preview
Qwen · Mungert
Params30.5B
Context256K
Q4_K_M18.9 GB
Vision
HO
Holotron4 30B A3B
Holotron · mradermacher
Params31.6B
Context1024K
Q4_K_M24.5 GB
GGUF
LL
Evolved Llama3 8B
Llama · mradermacher
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
MU
multi verse model i1
Multi · mradermacher
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
AM
AmoreLLM 1.7B
Amorellm · mradermacher
Params1.72B
Context32K
Q4_K_M1.1 GB
ReasoningTool use
ME
Melinoe 30B A3B Thinking Diluted
Melinoe · mradermacher
Params30.5B
Context256K
Q4_K_M18.6 GB
Reasoning
OL
olmOCR 7B 0225 Preview
OLMo · mradermacher
Params7.62B
Context32K
Q4_K_M4.7 GB
GGUF
LL
Llama 3 ChocoLlama 8B instruct i1
Llama · mradermacher
Params8B
Context8K
Q4_K_M4.9 GB
GGUF
QW
Qwen3 8B Thinking RoleVerse SFT DPO
Qwen · mradermacher
Params8.19B
Context40K
Q4_K_M5.0 GB
ReasoningRoleplayTool use
CT
Cthulhu 8B v1.4
Cthulhu · mradermacher
Params8B
Context1048K
Q4_K_M4.9 GB
Roleplay
TO
Tongyi-DeepResearch-30B-A3B
Tongyi · Mungert
Params30.5B
Context128K
Q4_K_M19.0 GB
GGUF
QW
EgoActor 8b Qwen3VL
Qwen · mradermacher
Params8.19B
Context256K
Q4_K_M5.0 GB
GGUF
QW
Qwen2.5 Coder 1.5B Instruct
Qwen · duyntnet
Params1.54B
Context32K
Q4_K_M1.0 GB
Coding
CO
Codegemma 7b
Codegemma · mradermacher
Params8.54B
Context8K
Q4_K_M5.3 GB
GGUF
FO
Foundation Sec 1.1 8B Instruct
Foundation · mradermacher
Params8.03B
Context64K
Q4_K_M4.9 GB
GGUF
GL
Qwopus GLM 18B Healed
GLM · mradermacher
Params15.9B
Context256K
Q4_K_M9.8 GB
ReasoningTool use
GE
Gemma 4 26B A4B It Arli
Gemma · mradermacher
Params25.2B
Context256K
Q4_K_M16.8 GB
GGUF
BO
boto-27B
Boto · mradermacher
Params27.2B
Context8K
Q4_K_M16.6 GB
GGUF
FL
Aurostral Small
Flamguard · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
GGUF

Other quants

Keep browsing