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.
QW
Qwen2.5 7B Instruct Abliterated v2
Qwen · mradermacher
Params7.62B
Context32K
Best fitQ2_K 4.0 GB
UN
Undi95 Emerhyst 20B
Undi · RichardErkhov
Params920M
Context—
Best fitIQ3_S 1.0 GB
MI
Dolphin Mistral 24B Venice Edition
Mistral · douganger
Params23.6B
Context128K
Q4_K_M14.3 GB
BG
bge-small-zh-v1.5
BGE · CompendiumLabs
Params24M
Context512
Best fitF32 0.7 GB
ME
Medgemma-1.5-4B-It
Medgemma · unsloth
Params3.88B
Context128K
Best fitIQ4_NL 4.0 GB
NE
privacy-filter-nemotron
Nemotron · LocalAI-io
Params1.4B
Context128K
Best fitF16 3.4 GB
RO
Rocinante X 12B v1
Rocinante · bartowski
Params12.2B
Context128K
Q4_K_M7.5 GB
NA
Nanbeige4.2 3B
Nanbeige · gkraker04
Params4.17B
Context256K
Best fitQ4_1 3.9 GB
JA
Jan code 4b
Jan · janhq
Params4.4B
Context256K
Best fitQ3_K_M 4.0 GB
GE
26B A4B
Gemma · terrorswift
Params25.2B
Context256K
Smallest0.5 GB
QW
Qwen3.8 Flash Next IQ3 S
Qwen · boxrunner
Params177B
Context256K
Best fitIQ4_NL 3.7 GB
MU
Muse Glimmer 30B Abliterated
Muse · The-freezer
Params28B
Context128K
Best fitQ4_K_M 2.3 GB
MI
MiniCPM V 4.7 1B DSpark
MiniCPM · openbmb
Params214M
Context256K
Best fitF16 1.0 GB
HO
K2 Horizon 0.9B
Horizon · IFM
Params1.08B
Context128K
Best fitBF16 3.1 GB
GL
Huihui GLM 4.7 Flash Abliterated
GLM · mradermacher
Params30B
Context198K
Q4_K_M18.1 GB
AP
35B A3B
Apus · apus-ailab
Params34.7B
Context256K
Q4_K_M21.2 GB
QW
Qwen3 Reranker 4B GGUF llama cpp
Qwen · Voodisss
Params4B
Context40K
Best fitQ3_K_M 3.8 GB
SM
Smollm3-3B
SmolLM · unsloth
Params3.08B
Context64K
Best fitQ6_K_XL 3.9 GB
QW
Qwen2 VL 2B Instruct
Qwen · ggml-org
Params1.54B
Context32K
Best fitF16 3.8 GB
LL
4yo1 llama3 pre1 pre2 ds ins2 lora3
Llama · RichardErkhov
Params200M
Context—
Best fitIQ3_XS 1.5 GB
QW
Qwen3 ASR 1.7B
Qwen · unslothai
Params2.03B
Context64K
Best fitQ8_0 3.6 GB
QW
Huihui Qwen3 4B Abliterated v2
Qwen · mradermacher
Params4B
Context40K
Best fitIQ4_XS 4.0 GB
GL
GLM 5.3 Flash UNCENSORED mlx oQ4e DFlash2
GLM · Solstice-AI
Params1.17B
Context1024K
Best fitBF16 3.0 GB
QW
Qwen3 VL 32B Instruct
Qwen · bartowski
Params32.8B
Context256K
Q4_K_M19.8 GB
MI
Ministral-3-8B-Reasoning-2512
Mistral · unsloth
Params8.49B
Context256K
Best fitIQ1_M 4.0 GB
MI
MiniCPM-V-4.7-1B
MiniCPM · AtomicChat
Params1.01B
Context256K
Best fitQ8_0 2.0 GB
QW
Qwen2.5 VL 7B Instruct Unredacted MAX
Qwen · mradermacher
Params7.62B
Context125K
Best fitQ2_K 4.0 GB
OC
Occamy 1.0 Abliterated
Occamy · SC117
Params34.7B
Context256K
Best fitBF16 1.4 GB
QW
Qwen3-4B-Thinking-2507
Qwen · unsloth
Params4B
Context256K
Best fitIQ4_XS 4.0 GB
PH
phonon 40m de
Phonon · gradium
Params51M
Context—
Smallest0.1 GB