GGUF models that fit in 4 GB

79024 GGUF language models have at least one quant whose estimated memory use is 4 GB or less at 8K context. Each card shows the largest quant that fits. Sizes run from 1M to 20260731B parameters. The longest trained context is 1280000K tokens.

Most downloadedNewestLargestLongest contextSmallest
XX
encoder
Xxl · chatpig
Params219M
Context512
Best fitF16 4.0 GB
Coding
OR
Ornith 1.5 9b Mtp
Ornith · protoLabsAI
Params456M
Context256K
Best fitQ8_0 4.0 GB
Vision
LY
Lythri 4B A2B
Lythri · bartowski
Params4.63B
Context128K
Best fitQ3_K_L 3.7 GB
GGUF
DE
WhiteRabbitNeo WhiteRabbitNeo V3 7B
Deephat · bartowski
Params7.62B
Context32K
Best fitQ2_K 4.0 GB
Coding
LF
LiquidAI_LFM2.5 1.2B Instruct
Lfm · lmstudio-community
Params1.17B
Context125K
Best fitQ8_0 2.8 GB
GGUF
GL
Glm-4-9B-0414
GLM · unsloth
Params9.4B
Context32K
Best fitIQ1_S 4.0 GB
GGUF
HO
K2-Horizon-MoVA-36B-A4B
Horizon · mradermacher
Params37.4B
Context512K
Q4_K_M22.4 GB
GGUF
MI
MiniCPM5 2B Math
MiniCPM · Caldalis
Params2.52B
Context128K
Best fitQ8_0 3.5 GB
Reasoning
NE
NVIDIA Nemotron 3.5 Bolt E4B Uncensored
Nemotron · AtomicChat
Params7.5B
Context64K
Q4_K_M5.1 GB
GGUF
MA
UI Mate 9B
Mate · bartowski
Params8.95B
Context256K
Q4_K_M5.9 GB
Vision
FA
Fara1.5 9B
Fara · bartowski
Params8.95B
Context256K
Q4_K_M5.9 GB
Vision
MI
MiniCPM V 4.7 1B heretic
MiniCPM · saidutta69
Params1.01B
Context256K
Best fitF16 2.9 GB
ReasoningRoleplayTool use
DA
DanTagGen beta
Dantaggen · KBlueLeaf
Params393M
Context512
Best fitF16 2.1 GB
GGUF
QW
Qwen3 VL 30B A3B Instruct
Qwen · bartowski
Params30.5B
Context256K
Q4_K_M18.6 GB
Vision
GE
Gemma 4 E2B it Artemis
Gemma · turintech
Params4.63B
Context128K
Best fitQ3_K_L 2.9 GB
ReasoningTool useVision
NE
Huihui NeoHorse 1 4B Abliterated MTP
Neohorse · huihui-ai
Params4.33B
Context256K
Best fitQ3_K 3.9 GB
ReasoningTool useVision
QW
Qwen3.8 9B Heretic Uncensored
Qwen · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
GGUF
OR
Orpheus-3B-0.1-Ft
Orpheus · unsloth
Params3.3B
Context128K
Best fitQ5_K_S 3.8 GB
GGUF
QW
Qwen3.8 27B Heretic JP Roleplay NSFW
Qwen · mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
Roleplay
QW
Qwen3 4B Thinking 2507
Qwen · lmstudio-community
Params4B
Context256K
Best fitQ3_K_L 3.9 GB
Reasoning
MU
music flamingo
Music · henry1477
Params7.6B
Context32K
Best fitQ2_K 4.0 GB
ReasoningVision
LL
Llama 3.2 1B Instruct Uncensored
Llama · mradermacher
Params1.24B
Context128K
Best fitF16 3.2 GB
GGUF
QW
Qwen2.5 7B Instruct Uncensored
Qwen · QuantFactory
Params7.6B
Context32K
Best fitQ2_K 4.0 GB
GGUF
GL
Glm-Z1-9B-0414
GLM · unsloth
Params9.4B
Context32K
Best fitIQ1_S 4.0 GB
GGUF
QW
Qwen3.8 27B Fable Distill Heretic Ara
Qwen · mradermacher
Params27.3B
Context256K
Q4_K_M16.8 GB
GGUF
QW
Qwen3.8 27B EfficientThink VAL APEX I NanoPlus V3
Qwen · IsValorum
Params27.3B
Context256K
Best fitQ4_K_M 1.8 GB
CodingReasoningTool use
BE
Behemoth 128B v3
Behemoth · bartowski
Params125B
Context256K
Q4_K_M78.4 GB
GGUF
QW
Qwen2.5.1 Coder 7B Instruct
Qwen · bartowski
Params7.62B
Context32K
Best fitQ2_K 4.0 GB
Coding
DA
Dark Scarlett 27B v2.0
Dark · ReadyArt
Params27.3B
Context256K
Q4_K_M16.8 GB
RoleplayVision
AP
APUS Reflex 4B
Apus · apus-ailab
Params4.2B
Context256K
Q4_K_M2.7 GB
GGUF

Other memory sizes

Keep browsing