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.
SK
Sketch Cydonia 24B V1.3
Sketch · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
JI
jina embeddings v4 text retrieval
Jina · jinaai
Params3.1B
Context125K
Best fitQ6_K 3.3 GB
MU
Muse Glimmer 30B
Muse · AaryanK
Params28B
Context128K
Best fitQ8_0 3.7 GB
SL
Slimaki Tavern 24B v1.3
Slimaki · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB
TO
Tongyi DeepResearch 30B A3B
Tongyi · bartowski
Params30.5B
Context128K
Q4_K_M18.6 GB
GE
Gemma 3 1b
Gemma · lmstudio-community
Params1B
Context32K
Best fitQ4_0 1.4 GB
FA
90m Instruct
Falcon · tiiuae
Params91M
Context256K
Best fitBF16 0.8 GB
SE
Seed-Coder-8B-Reasoning
Seed · unsloth
Params8.25B
Context64K
Best fitIQ1_M 4.0 GB
QW
Qwen3.5 9B Heretic Ara v1
Qwen · mradermacher
Params8.95B
Context256K
Q4_K_M5.6 GB
QW
Qwen3.5 35B A3B Heretic
Qwen · mradermacher
Params34.7B
Context256K
Q4_K_M21.2 GB
GE
Gemma 3 1b It Abliterated
Gemma · bartowski
Params1B
Context32K
Best fitBF16 2.7 GB
GE
Baihua Gemma4 E4B
Gemma · mradermacher
Params7.46B
Context128K
Q4_K_M5.3 GB
D1
d1 3B
NANI-Nithin
Params2.7B
Context32K
Best fitQ3_K_L 4.0 GB
GR
Granite 4.0 1b
Granite · ibm-granite
Params1.63B
Context128K
Best fitQ3_K_S 4.0 GB
AR
Arctic Text2SQL R1 7B
Arctic · mradermacher
Params7.62B
Context32K
Best fitQ2_K 4.0 GB
GR
Granite 8b Code Base 128k
Granite · mradermacher
Params8.05B
Context125K
Best fitIQ2_XXS 3.9 GB
GE
Gemma 3 270m It
Gemma · ggml-org
Params268M
Context32K
Best fitQ8_0 0.9 GB
WH
whisper tiny
Whisper · oxide-lab
Params38M
Context—
Best fitQ2_K 0.5 GB
QW
Huihui Qwen3 VL 4B Instruct Abliterated
Qwen · mradermacher
Params4B
Context256K
Best fitIQ4_XS 4.0 GB
NE
nexus medical
Nexus · King3Djbl
Params1.54B
Context32K
Best fitF16 3.8 GB
MI
Mistral Nemo 2407 12B Thinking Claude Gemini GPT5.2 Uncensored HERETIC
Mistral · mradermacher
Params12.2B
Context1000K
Q4_K_M7.5 GB
SP
Spark X2.6 1.7B
Spark · XHToken
Params1.98B
Context1024K
Best fitQ8_0 2.8 GB
MI
MiniCPM5 1B
MiniCPM · NANI-Nithin
Params1.08B
Context128K
Best fitBF16 2.9 GB
QW
Qwen_Qwen3 0.6B
Qwen · lmstudio-community
Params752M
Context40K
Best fitQ8_0 2.2 GB
QW
Qwen3-Coder-Next
Qwen · Mungert
Params79.7B
Context256K
Q4_K_M49.2 GB
DE
DeepScaleR 1.5B Preview
Deepscaler · bartowski
Params1.78B
Context128K
Best fitQ8_0 2.6 GB
QW
Qwen3 VL 8B Abliterated Caption It
Qwen · prithivMLmods
Params8.19B
Context256K
Best fitIQ1_M 4.0 GB
QW
Qwen2.5 Coder 7B
Qwen · bartowski
Params7.62B
Context32K
Best fitQ2_K 4.0 GB
LL
Llama3-Chinese-8B-Instruct
Llama · RichardErkhov
Params8B
Context8K
Best fitIQ3_S 3.3 GB
MA
Magidonia 24B v4.3 Heretic v1.2
Magidonia · mradermacher
Params23.6B
Context128K
Q4_K_M14.3 GB