Tiny GGUF models
62726 GGUF models with 3B parameters or fewer. Sizes run from 1M to 3.5B parameters. The longest trained context is 795898K tokens.
BO
Borealis 1b
Borealis · NbAiLab
Params1B
Context32K
Q4_K_M0.8 GB
QW
Qwen3.5 0.8B GPT6 Luna Distill Karpy
Qwen · NasledieLab
Params752M
Context256K
Q4_K_M0.5 GB
QW
Qwen3.8 Flash Next MTP block
Qwen · paperniuk
Params2.61B
Context256K
Smallest1.5 GB
KH
KhanomTanLLM Instruct
Khanomtanllm · mradermacher
Params3.45B
Context2K
Q4_K_M2.6 GB
TA
UI TARS 2B SFT
Tars · lmstudio-community
Params1.78B
Context32K
Q4_K_M1.1 GB
QW
Qwen3-ASR-1.7B
Qwen · sakasegawa
Params2.04B
Context—
Smallest1.2 GB
PY
Pythia 2.8b
Pythia · mradermacher
Params2.78B
Context2K
Q4_K_M1.8 GB
QW
Qwen3.8 27B DFlash2 Bonsai2 ternary
Qwen · Cyronius
Params1.92B
Context256K
Smallest0.7 GB
RH
rho math 1b interpreter v0.1
Rho · QuantFactory
Params1.1B
Context2K
Q4_K_M0.7 GB
LL
Llama3.2 3B ShiningValiant2
Llama · QuantFactory
Params3.21B
Context128K
Q4_K_M2.0 GB
DE
deepseek-coder-1.3B-kexer
DeepSeek · bartowski
Params1.35B
Context16K
Q4_K_M0.9 GB
PA
palmyra mini thinking AIO
Palmyra · prithivMLmods
Params1.78B
Context128K
Q4_K_M1.0 GB
QW
Qwen3.8 27B
Qwen · Brian6145
Params461M
Context—
Q4_K_M16.8 GB
MI
Ministral 3 3B Instruct 2512 Absolute Heresy
Mistral · mradermacher
Params3.43B
Context256K
Q4_K_M2.1 GB
MA
ms marco MiniLM L 6 v2
Marco · cstr
Params23M
Context512
Smallest0.0 GB
QW
Indian Legal Qwen2.5 3B
Qwen · GSMS-B
Params3.09B
Context32K
Q4_K_M1.9 GB
QW
Qwen2.5 Coder 1.5B Instruct non censure uncensored coder for android
Qwen · Tree-AI-lab
Params1.78B
Context32K
Q4_K_M1.1 GB
GR
Granite Speech 5.0 2b
Granite · mradermacher
Params1.84B
Context16K
Q4_K_M1.1 GB
NE
NVIDIA Nemotron 3.5 Lightning 30B A3B NVFP4 DFlash
Nemotron · magnitudedev
Params833M
Context1024K
Smallest1.2 GB
OP
knoveleng Open RS3
Open · bartowski
Params1.78B
Context128K
Q4_K_M1.1 GB
SA
Sailor-0.5B
Sailor · mradermacher
Params620M
Context32K
Q4_K_M0.4 GB
QW
Qwen2.5 VL 3B Instruct heretic
Qwen · saidutta69
Params3.09B
Context125K
Q4_K_M1.9 GB
BI
Bielik-1.5B-v3.0-Instruct
Bielik · speakleash
Params1.6B
Context8K
Smallest1.7 GB
QW
Qwen3-Reranker-0.6B
Qwen · Mungert
Params596M
Context40K
Q4_K_M0.4 GB
MI
Ministral 3 3B Instruct 2512 roleplay tuned ERP tolerant
Mistral · Indexnusrefather
Params3.43B
Context256K
Q4_K_M2.1 GB
FL
Florence 2 base
Florence · antflydb
Params232M
Context—
Smallest0.1 GB
PY
Pythia 410m Deduped
Pythia · mradermacher
Params405M
Context2K
Q4_K_M0.3 GB
OP
OpenELM-450M
Openelm · RichardErkhov
Params457M
Context2K
Q4_K_M0.3 GB
MA
maya1
Maya · mradermacher
Params3.3B
Context128K
Q4_K_M2.1 GB
GR
Granite Speech 5.0 470M TurboCTC
Granite · handy-computer
Params473M
Context—
Q4_K_M0.3 GB