GGUF models that fit in 48 GB

201364 GGUF language models have at least one quant whose estimated memory use is 48 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.

Most downloadedNewestLargestLongest contextSmallest
QW
Qwen2.5 VL 7B Instruct
Qwen · ggml-org
Params7.62B
Context125K
Best fitF16 16.2 GB
Vision
LL
Meta Llama 3.1 8B Instruct
Llama · lmstudio-community
Params8B
Context128K
Best fitQ8_0 10.1 GB
GGUF
QW
Qwen2.5 3B Instruct
Qwen · bartowski
Params3.09B
Context32K
Best fitF32 13.2 GB
GGUF
QW
qmd query expansion 1.7B
Qwen · tobil
Params2.03B
Context40K
Best fitF16 5.5 GB
GGUF
DE
DeepSeek R1 Distill Qwen 32B
DeepSeek · bartowski
Params32.8B
Context128K
Best fitQ8_0 37.5 GB
Reasoning
TO
Topxtral-4x7B-v0.1
Topxtral · RichardErkhov
Params18.5B
Context32K
Best fitQ4_K 12.8 GB
GGUF
QW
Qwen3 4B Instruct 2507
Qwen · bartowski
Params4B
Context256K
Best fitBF16 9.8 GB
GGUF
LI
Ling 3.0 Flash Dspark
Ling · bloomer010
Params127B
Context256K
Best fitQ2_K_XL 44.6 GB
GGUF
TI
tinygemma3
Tinygemma · ggml-org
Params38M
Context128K
Best fitQ8_0 0.7 GB
Vision
MO
moondream2 multiformat
Moondream · Serkan007
Params—
Context—
Best fitQ4_K_M 0.5 GB
Coding
MI
Hermes 2 Pro Mistral 7B
Mistral · NousResearch
Params7.24B
Context32K
Best fitQ8_0 9.3 GB
Tool use
QW
Qwen3 30B A3B Instruct 2507
Qwen · bartowski
Params30.5B
Context256K
Best fitQ8_0 33.8 GB
GGUF
QW
Qwen3.6 40B Claude 4.6 Opus Deckard Heretic Uncensored Thinking NEO CODE Di IMatrix MAX
Qwen · DavidAU
Params39.1B
Context256K
Best fitQ8_0 46.6 GB
CodingReasoningRoleplay
QW
Qwen3-Next-80B-A3B-Instruct
Qwen · unsloth
Params80B
Context256K
Best fitQ4_K_XL 47.4 GB
GGUF
DE
DeepSeek R1 Distill Qwen 7B
DeepSeek · unsloth
Params7.62B
Context128K
Best fitF16 16.2 GB
Reasoning
DE
DeepSeek R1 Distill Qwen 14B
DeepSeek · bartowski
Params14.8B
Context128K
Best fitF16 31.7 GB
Reasoning
XI
Xing4.0 29B A4B
Xing · Venastine-Research
Params31.2B
Context256K
Best fitQ8_0 34.5 GB
Coding
MI
Mistral-Small-3.2-24B-Instruct-2506
Mistral · unsloth
Params23.6B
Context128K
Best fitQ8_K_XL 30.8 GB
Vision
OP
openthaigpt1.5 7b instruct
Openthaigpt · openthaigpt
Params7.62B
Context32K
Best fitF16 16.2 GB
GGUF
QW
Models Qwen Qwen3 0.6B
Qwen · MaziyarPanahi
Params752M
Context40K
Q4_K_M0.5 GB
GGUF
QW
Qwen3.5-122B-A10B
Qwen · unsloth
Params125B
Context256K
Best fitQ2_K_XL 44.2 GB
Vision
QW
Qwen3.8 27B
Qwen · AtomicChat
Params27.3B
Context256K
Best fitQ8_0 31.6 GB
ReasoningTool useVision
NO
nomic-embed-text-v2-moe
Nomic · nomic-ai
Params475M
Context512
Best fitF32 2.7 GB
Embedding
QW
Qwen3Vl 30b A3B Instruct
Qwen · Qwen
Params30.5B
Context256K
Best fitQ8_0 33.8 GB
Vision
QW
Qwen3-14B
Qwen · unsloth
Params14.8B
Context40K
Best fitBF16 31.4 GB
GGUF
TI
tinyllama_tinyllama-1.1b-chat-v1.0
Tinyllama · TheBloke
Params1.1B
Context2K
Best fitQ8_0 1.9 GB
Coding
QW
Models Qwen Qwen3 4B
Qwen · MaziyarPanahi
Params4B
Context40K
Q4_K_M2.5 GB
GGUF
SM
SmolLM2 135M Instruct
SmolLM · unsloth
Params135M
Context8K
Best fitF16 1.0 GB
GGUF
LL
Llama 4 Scout 17B 16E Instruct
Llama · bartowski
Params108B
Context10240K
Best fitIQ3_XXS 47.1 GB
GGUF
GE
gemma 4 E2B it qat
Gemma · unsloth
Params4.63B
Context128K
Best fitQ4_K_XL 3.2 GB
Vision

Other memory sizes

Keep browsing