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.
GE
Winnow E2B
Gemma · EldanRing
Params77M
Context128K
Best fitBF16 0.7 GB
CH
Chaotic Order V4
Chaotic · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
HO
K2 Horizon 3.7B Claude Opus Reasoning Distilled
Horizon · mradermacher
Params5.06B
Context512K
Best fitQ2_K 3.9 GB
GO
Goetia 24B v1.4 Heretic
Goetia · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
TO
ToolSpace 35B A3B
Toolspace · mradermacher
Params34.7B
Context256K
Q4_K_M21.2 GB
GE
Gemma 2 2b It
Gemma · l3utterfly
Params2.61B
Context8K
Best fitQ6_K 3.5 GB
QW
Qwen2.5 3B
Qwen · mradermacher
Params3.09B
Context32K
Best fitQ6_K 3.3 GB
QW
ColdBrew Qwen3.5 RP_Base CharThink
Qwen · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
MI
YuisekinAIEvol Mistral 7B ja math v0.1.1
Mistral · mmnga
Params7.24B
Context32K
Best fitIQ2_S 3.9 GB
IN
InternVL3_5 4B
Internvl · bartowski
Params4.41B
Context40K
Best fitQ3_K_M 4.0 GB
SN
Snowpiercer 15B v4
Snowpiercer · bartowski
Params15B
Context64K
Q4_K_M9.1 GB
WO
WorldSim Opus 3.6 35B A3B
Worldsim · bartowski
Params34.7B
Context256K
Q4_K_M21.4 GB
EX
EXAONE 3.5 7.8B Instruct
Exaone · ThomasBaruzier
Params7.8B
Context32K
Best fitIQ2_XXS 3.9 GB
QW
ThinCtx Qwen3.5 4B
Qwen · mradermacher
Params4.21B
Context256K
Best fitQ3_K_L 4.0 GB
D3
d3 i1
mradermacher
Params27B
Context256K
Q4_K_M16.5 GB
LF
LFM2.5 2.6B Uncensored Trial65
Lfm · SC117
Params2.7B
Context125K
Best fitIQ4_XS 4.0 GB
LF
LFM3-Omni-2B
Lfm · AtomicChat
Params1.73B
Context125K
Best fitQ8_0 4.0 GB
MI
MiMo V2.6 Flash REAP50
Mimo · patrickbdevaney
Params157B
Context1024K
Best fitQ8_0 3.9 GB
FR
FrogNano 4B 2609 Heretic
Frognano · mradermacher
Params4.21B
Context256K
Best fitQ3_K_L 4.0 GB
NO
Noema 1.5 4B Deep
Noema · mradermacher
Params4.21B
Context256K
Best fitQ3_K_L 4.0 GB
QW
Qwen3.5 4B Anon Reasoning v2
Qwen · mradermacher
Params4.21B
Context256K
Best fitQ3_K_L 4.0 GB
OR
Ornith 1.0 35B Uncensored Heretic
Ornith · mradermacher
Params34.7B
Context256K
Q4_K_M21.2 GB
PY
Pythia 160m Deduped
Pythia · mradermacher
Params162M
Context2K
Best fitIQ1_S 0.9 GB
SU
SummLlama3.1 8B
Summllama · mradermacher
Params8.03B
Context128K
Best fitIQ2_XXS 4.0 GB
LL
Llama 3.2 1B Instruct Abliterated
Llama · mradermacher
Params1.5B
Context128K
Best fitQ6_K 2.0 GB
FI
Fin R1
Fin · mradermacher
Params7.6B
Context32K
Best fitQ2_K 4.0 GB
AP
Apertus 70B Instruct 2509 Abliterated
Apertus · mradermacher
Params70.6B
Context64K
Q4_K_M43.7 GB
QW
Huihui Qwen3.5 4B Claude 4.6 Opus Abliterated
Qwen · mradermacher
Params4.21B
Context256K
Best fitQ3_K_L 4.0 GB
PH
Phi 4 Reasoning plus_Nemotron Reasoning 0803_SFT GRPO
Phi · mradermacher
Params14.7B
Context32K
Q4_K_M9.1 GB
IN
InternLM2
InternLM · legraphista
Params1.89B
Context8K
Best fitQ8_0 3.3 GB