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.
QW
Qwen2.5-Vl-32B-Instruct
Qwen · unsloth
Params32.8B
Context125K
Best fitIQ2_XXS 11.9 GB
SH
Shieldgemma 2b
Shieldgemma · legraphista
Params2.61B
Context8K
Best fitBF16 6.6 GB
QW
Qwen3.6 27B ABLITERATED UNCENSORED PHILADELPHIA CLASS
Qwen · mradermacher
Params26.9B
Context256K
Best fitIQ2_S 12.0 GB
QW
dolphin-2.9.3-qwen2-1.5b
Qwen · crusoeai
Params1.54B
Context128K
Q4_K_M1.0 GB
AM
Amaretto 8B
Amaretto · mradermacher
Params8.49B
Context256K
Best fitQ6_K 8.6 GB
QW
Josiefied Qwen3 14B Abliterated v3
Qwen · bartowski
Params14.8B
Context40K
Best fitQ4_K_L 11.4 GB
PH
phi 4 i1
Phi · mradermacher
Params14.7B
Context16K
Best fitQ4_1 11.4 GB
MI
Pixtral
Mistral · bartowski
Params12.2B
Context1000K
Best fitQ6_K 11.9 GB
TE
tenyx_tenyxchat-7b-v1
Tenyxchat · TheBloke
Params7.24B
Context8K
Best fitQ8_0 9.3 GB
PY
Pythia 6.9b Deduped
Pythia · mradermacher
Params6.86B
Context2K
Best fitQ6_K 10.4 GB
HA
Haru Kanana 8b
Haru · mradermacher
Params8B
Context32K
Best fitQ6_K 8.2 GB
LF
LFM2.5-24B-A2B
Lfm · mudler
Params24B
Context128K
Smallest7.7 GB
YI
Yi-1.5-9B-Chat-16K
Yi · mradermacher
Params8.83B
Context16K
Best fitQ6_K 8.6 GB
SM
nisten/Biggie-SmoLlm-0.15B-Base
SmolLM · nisten
Params152M
Context2K
Best fitBF16 1.1 GB
QW
QiMing Holos Plus Qwen3 14B
Qwen · mradermacher
Params14.8B
Context40K
Best fitQ4_1 11.2 GB
UN
Hy MT2 30B A3B Uncensored Heretic
Uncensored · alphaZimuth
Params30B
Context256K
Smallest9.7 GB
ST
L3-8B-Stheno-v3.2
Stheno · mradermacher
Params8B
Context8K
Best fitQ6_K 8.2 GB
MI
meta-math_metamath-mistral-7b
Mistral · TheBloke
Params7.24B
Context32K
Best fitQ8_0 9.3 GB
ZA
Zarafusionex 1.1 L2 7B
Zarafusionex · TheBloke
Params6.74B
Context4K
Best fitQ8_0 12.0 GB
MI
dolphin-2.9.1-mixtral-1x22b
Mistral · mradermacher
Params22.2B
Context64K
Best fitQ3_K_S 12.0 GB
QW
Jan-v2-VL-high
Qwen · Mungert
Params8.2B
Context256K
Best fitQ8_0 10.4 GB
QW
Qwen3.6 35B A3B Uncensored Aggressive
Qwen · mradermacher
Params34.7B
Context256K
Best fitIQ2_S 11.8 GB
QW
Qwen3 VL 32B Instruct Ultra Uncensored Heretic
Qwen · mradermacher
Params32.8B
Context256K
Best fitIQ2_XXS 11.7 GB
HO
Holo4 27B Abliterix
Holo · mradermacher
Params27B
Context256K
Best fitIQ2_S 12.0 GB
QW
Qwen2.5 Coder 0.5B Instruct Abliterated
Qwen · bartowski
Params494M
Context32K
Best fitF16 1.6 GB
MI
Mistral Nemo 12B Instruct
Mistral · mradermacher
Params12.2B
Context1000K
Best fitQ6_K 11.9 GB
TO
TowerInstruct 7B v0.1
Towerinstruct · TheBloke
Params6.74B
Context4K
Best fitQ8_0 12.0 GB
QW
Qwen3 8B Heretic
Qwen · mradermacher
Params8.19B
Context40K
Best fitQ6_K 8.4 GB
GE
Gemma 3 it 4B Uncensored DBL X
Gemma · DavidAU
Params4.26B
Context128K
Best fitF16 10.3 GB
GP
GPT OSS Cybersecurity 20B Merged Heretic
gpt-oss · mradermacher
Params20.9B
Context128K
Q4_K_M15.8 GB