DeepSeek GGUF models

4419 GGUF models in the DeepSeek family, quantized by 1572 uploaders. Sizes run from 1M to 5376B parameters. The longest trained context is 1024K tokens.

TinySmallMediumLarge
Most downloadedNewestLargestLongest contextSmallest
DE
DeepSeek R1 Distill Llama 8B NexaQuant
DeepSeek · NexaAI
Params8.03B
Context128K
Smallest5.3 GB
ReasoningTool use
DE
DeepSeek GRM 27B
DeepSeek · mradermacher
Params27.2B
Context8K
Q4_K_M16.6 GB
Tool use
DE
DeepSeek R1 Refined2 Llama 3.1 8B hf
DeepSeek · mradermacher
Params8.03B
Context128K
Q4_K_M4.9 GB
ReasoningTool use
DE
Deepseek R1 Distill Llama 8B
DeepSeek · steampunque
Params8.03B
Context128K
Smallest6.5 GB
ReasoningTool use
DE
DeepSeek R1 ReDistill Llama3 8B v1.1
DeepSeek · mradermacher
Params8.03B
Context128K
Q4_K_M4.9 GB
ReasoningTool use
DE
deepseek deepseek v4 1 6t pro
DeepSeek · cPilotGod
Params1570B
Context1024K
Q4_K_M954.6 GB
ReasoningTool use
DE
FuseO1 DeepSeekR1 Qwen2.5 Coder 32B Preview
DeepSeek · DataSoul
Params32.8B
Context128K
Q4_K_M19.9 GB
CodingTool use
DE
LF CODER DEEPSEEK1.3
DeepSeek · mradermacher
Params1.35B
Context16K
Q4_K_M0.9 GB
Coding
DE
DeepSeek V4.1 Flash MXFP4 Q8 0 experimental
DeepSeek · mussonking
Params748B
Context1024K
Smallest508.0 GB
GGUF
DE
DeepSeek V4.1 Flash 1017 GSQ 217 Strixolotl
DeepSeek · Snowsh1ne
Params14.2B
Context—
Smallest5.9 GB
GGUF
DE
DeepSeek V4 Flash 0731
DeepSeek · Stk88
Params284B
Context1024K
Smallest10.9 GB
GGUF
DE
DeepSeek V2 Lite XiaAi
DeepSeek · mradermacher
Params15.7B
Context160K
Q4_K_M10.4 GB
GGUF
DE
deepseek deepseek r1 distill 70b
DeepSeek · cPilotGod
Params70.6B
Context128K
Q4_K_M42.5 GB
ReasoningTool use
DE
DeepSeek R1 Distill Qwen 14B
DeepSeek · Fastiraz
Params14.8B
Context128K
Q4_K_M9.0 GB
ReasoningTool use
DE
muse deepseek7b
DeepSeek · MCult01
Params6.91B
Context4K
Q4_K_M4.2 GB
GGUF
DE
DeepSeek Coder V2 Lite Instruct
DeepSeek · legionarius
Params15.7B
Context160K
Q4_K_M10.4 GB
Coding
DE
DeepSeek Coder V2 Lite Base
DeepSeek · alphaduriendur
Params15.7B
Context160K
Q4_K_M10.4 GB
Coding
DE
Qwen2.5 1.5B DeepSeek R1 Instruct
DeepSeek · mradermacher
Params1.78B
Context32K
Q4_K_M1.1 GB
ReasoningTool use
DE
Deepseek Coder 1.3b Instruct
DeepSeek · alphaduriendur
Params1.35B
Context16K
Q4_K_M0.9 GB
Coding
DE
Qwen2.5 32B DeepSeek R1 Instruct
DeepSeek · mradermacher
Params32.8B
Context32K
Q4_K_M19.9 GB
ReasoningTool use
DE
Huihui ai.DeepSeek R1 Distill Llama 70B Abliterated
DeepSeek · DevQuasar-7
Params70.6B
Context128K
Q4_K_M42.5 GB
ReasoningTool use
DE
Deepseek_1.5_Reddit_Left
DeepSeek · mradermacher
Params1.78B
Context128K
Q4_K_M1.1 GB
ReasoningTool use
DE
NEW Fuse DeepSeek R1 AIME 1
DeepSeek · mradermacher
Params32.8B
Context128K
Q4_K_M19.9 GB
ReasoningTool use
DE
Lumo DeepSeek R1 8B
DeepSeek · lumolabs-ai
Params8.03B
Context128K
Smallest4.7 GB
ReasoningTool use
DE
DeepSeek-R1-Distill-Qwen-1.5B-C
DeepSeek · mradermacher
Params1.78B
Context128K
Q4_K_M1.1 GB
ReasoningTool use
DE
LFM2.5 1.2B MEGABRAIN Thinking Claude Polaris Deepseek GLM
DeepSeek · LoveGlaze
Params1.17B
Context125K
Q4_K_M0.7 GB
Reasoning
DE
FuseO1 DeepSeekR1 QwQ SkyT1 32B Preview
DeepSeek · FuseAI
Params32.8B
Context128K
Q4_K_M19.9 GB
ReasoningTool use
DE
DeepSeek R1 Distill Qwen 14B Blunt
DeepSeek · mradermacher
Params14.8B
Context128K
Q4_K_M9.0 GB
ReasoningTool use
DE
DeepSeek R1 Distill Qwen 32B
DeepSeek · 1bit-MONSTER
Params32.8B
Context128K
Q4_K_M19.9 GB
ReasoningTool use
DE
DeepSeek V4 Flash 0731
DeepSeek · lyssquant
Params284B
Context1024K
Smallest10.9 GB
GGUF

Keep browsing