GGUF models that fit in 12 GB
171517 GGUF language models have at least one quant whose estimated memory use is 12 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 4194304K tokens.
MU
Muse-Glimmer-30B-Hermes-Agentic
Muse · vcruz305
Params27.9B
Context128K
Best fitQ2_K 11.6 GB
FL
flint
Flint · terelabs
Params39.8B
Context256K
Q4_K_M25.0 GB
QW
Qwen3.8 27B Humanlike Chat 2.0
Qwen · mradermacher
Params27.3B
Context256K
Q4_K_M16.8 GB
DE
Qwen3.5 9B Distill Deepseek V4.1 Flash 20261003
DeepSeek · kevinzhou21
Params9.2B
Context256K
Best fitQ8_0 11.4 GB
QW
Qwen3.8 27B Uncensored
Qwen · jaromer
Params27.3B
Context256K
Best fitQ8_0 5.8 GB
QW
Qwen3.8 9B Distill SLERP F451 Pro Writer Uncensored
Qwen · mradermacher
Params9.2B
Context256K
Best fitQ8_0 11.4 GB
CL
Clef-Flash
Clef · turintech
Params9.08B
Context256K
Best fitIQ4_XS 6.5 GB
HU
Huihui Ornith 1.5 9B Abliterated
Huihui · mradermacher
Params8.95B
Context256K
Best fitQ6_K 8.9 GB
PY
Pythia 2.8b Deduped
Pythia · mradermacher
Params2.78B
Context2K
Best fitQ6_K 5.5 GB
SM
SmolLM2 1.7B Instruct
SmolLM · lmstudio-community
Params1.71B
Context8K
Best fitQ8_0 3.9 GB
GR
Granite 34b Code Base 8k
Granite · mradermacher
Params33.7B
Context8K
Best fitIQ2_S 11.6 GB
LL
Llama 3.2 3B Instruct
Llama · DavidAU
Params3.21B
Context128K
Best fitQ6_K 4.1 GB
QW
MiroThinker-v1.5-30B
Qwen · Mungert
Params30.5B
Context256K
Best fitIQ2_XXS 11.7 GB
MI
Mahou-1.3a-mistral-7B
Mistral · mradermacher
Params7.24B
Context32K
Best fitQ6_K 7.5 GB
GR
Granite 3b Code Instruct 2k
Granite · mradermacher
Params3.48B
Context2K
Best fitQ6_K 6.0 GB
FO
foundational models
Foundational · sagalore-softworks
Params11.9B
Context256K
Smallest1.1 GB
QW
Qwen3 30B A3B Architect18
Qwen · mradermacher
Params30.5B
Context256K
Best fitQ2_K_S 11.8 GB
MI
Mixtral-8x7B-Instruct-v0.1-LimaRP-ZLoss
Mistral · mradermacher
Params46.7B
Context32K
Best fitIQ1_S 11.4 GB
HY
Hy MT2 1.8B 2Bit
tencent
Params1.79B
Context256K
Smallest0.6 GB
UM
Umbra-v3-MoE-4x11b-2ex
Umbra · mradermacher
Params36.1B
Context32K
Best fitIQ2_XXS 11.6 GB
PA
MS3.2 PaintedFantasy v4 24B
Paintedfantasy · bartowski
Params23.6B
Context128K
Best fitIQ3_XS 11.7 GB
MI
Roleplay-Mistral-7B
Mistral · mradermacher
Params7.24B
Context32K
Best fitQ6_K 7.5 GB
QW
flux2 klein 4B uncensored text encoder
Qwen · WeReCooking
Params4B
Context40K
Best fitQ4_0 4.1 GB
KA
KAT Coder V2.5 Dev 35B A3B MTP ABLITERATED
Kat · mradermacher
Params35.5B
Context256K
Q4_K_M21.7 GB
GE
Gemma 4 26b A4B Heretic Styletune v2 Head
Gemma · mradermacher
Params26B
Context256K
Q4_K_M17.2 GB
FI
Final Order 24B
Final · mradermacher
Params23.6B
Context128K
Best fitIQ3_XS 11.7 GB
JI
jina embeddings v4 text matching
Jina · jinaai
Params3.1B
Context125K
Best fitF16 7.0 GB
MI
cognitivecomputations_dolphin-2.6-mistral-7b-dpo-laser
Mistral · TheBloke
Params7.24B
Context32K
Best fitQ8_0 9.3 GB
HU
Huihui Hunyuan MT 7B Abliterated
Huihui · mradermacher
Params7.5B
Context256K
Best fitQ8_0 9.6 GB
QW
Qwen3 4B Hivemind Instruct Heretic Abliterated Uncensored NEO Imatrix
Qwen · DavidAU
Params4B
Context256K
Best fitQ8_0 6.0 GB