LLMs ·

DeepSeek V3 (671B)

Parameters
671B (37.0B active)
Architecture
Mixture of experts
Max context
128K tokens
License
deepseek
Released
2024-12

How to run it

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

ollama run deepseek-v3:671b

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/unsloth/DeepSeek-V3-GGUF (opens in a new tab)

ollama run hf.co/unsloth/DeepSeek-V3-GGUF

Original weights · Hugging Face

Full-precision files from DeepSeek, for fine-tuning, vLLM, Transformers.

huggingface.co/deepseek-ai/DeepSeek-V3 (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from DeepSeek. The weights themselves download from Hugging Face.

github.com/deepseek-ai/DeepSeek-V3 (opens in a new tab)

git clone https://github.com/deepseek-ai/DeepSeek-V3

Links checked 2 days ago against Hugging Face, Ollama and GitHub.

Apps that run DeepSeek V3 (671B)

Our pick for each kind of computer, assuming the model fits in its memory. Pick your own GPU or device in the Model Finder for an exact answer.

Mac (Apple silicon)

LM Studio

Runs the MLX build, typically up to 10% faster than GGUF for a model with 37B active parameters on Apple silicon.

Also good:

Windows PC, NVIDIA GPU

Ollama

One command with Ollama’s own build of this model.

Also good:

Windows PC, AMD Radeon

Lemonade

AMD’s own app: llama.cpp tuned for Radeon on Vulkan and ROCm.

Also good:

Linux, NVIDIA GPU

Ollama

One command with Ollama’s own build of this model.

Also good:

Linux server, 80 GB+ GPUs

Ollama

One command with Ollama’s own build of this model.

Also good:

App Type Opens it as Systems Engine Start
LM Studio Desktop app GGUF or MLX macOSWindowsLinux llama.cpp, MLX Open in LM Studio
Ollama CLI + API Ollama library macOSWindowsLinux llama.cpp, MLX 
ollama run deepseek-v3:671b
llama.cpp CLI + API GGUF macOSWindowsLinux llama.cpp 
llama-server -hf unsloth/DeepSeek-V3-GGUF:Q4_K_M
Jan Desktop app GGUF or MLX macOSWindowsLinux llama.cpp, MLX Open in Jan
MLX LM CLI + API MLX macOS MLX 
mlx_lm.chat --model mlx-community/DeepSeek-V3-4bit
vLLM Server Original weights WSLWindowsLinux vLLM 
vllm serve deepseek-ai/DeepSeek-V3
SGLang Server Original weights WSLWindowsLinux SGLang 
python3 -m sglang.launch_server --model-path deepseek-ai/DeepSeek-V3
Docker Model Runner CLI + API GGUF macOSWindowsLinux llama.cpp, vLLM 
docker model run hf.co/unsloth/DeepSeek-V3-GGUF:Q4_K_M
Lemonade Desktop app GGUF macOSWindowsLinux llama.cpp, ONNX Runtime GenAI 
lemonade pull unsloth/DeepSeek-V3-GGUF:Q4_K_M
Unsloth Desktop Desktop app GGUF or MLX macOSWindowsLinux llama.cpp, MLX, stable-diffusion.cpp Open in Unsloth Desktop
Atomic Chat Desktop app GGUF or MLX macOSWindowsLinux llama.cpp, MLX-VLM Open in Atomic Chat
GhidrAssist RE plugin GGUF or MLX macOSWindowsLinux uses Ollama, LM Studio or any OpenAI-compatible server Add it to Ghidra, then point it at a local model server
Gepetto RE plugin GGUF or MLX macOSWindowsLinux uses Ollama or an OpenAI-compatible server Add it to IDA Pro, then point it at a local model server
BinAssist RE plugin GGUF or MLX macOSWindowsLinux uses Ollama, LM Studio or any OpenAI-compatible server Add it to Binary Ninja, then point it at a local model server
r2ai and decai RE plugin GGUF or MLX macOSWindowsLinux uses Ollama, llama.cpp or LM Studio Add it to radare2, then point it at a local model server
GPT4All Desktop app GGUF macOSWindowsLinux llama.cpp Get GPT4All
KoboldCpp Desktop app GGUF macOSWindowsLinux llama.cpp Get KoboldCpp
Text Generation WebUI Desktop app GGUF macOSWindowsLinux llama.cpp, Transformers, ExLlamaV3 Get Text Generation WebUI
llamafile CLI + API GGUF macOSWindowsLinux llama.cpp Get llamafile
Msty Studio Desktop app GGUF or MLX macOSWindowsLinux Ollama, llama.cpp, MLX Open in Msty Studio
Llama (macOS) Desktop app GGUF macOS llama.cpp Open in Llama (macOS)
RecurseChat Desktop app GGUF macOS llama.cpp Open in RecurseChat
Inferencer Desktop app MLX macOS MLX Get Inferencer
node-llama-cpp Library GGUF macOSWindowsLinux llama.cpp Get node-llama-cpp
exo Server MLX macOSLinux MLX Get exo
PocketPal AI Mobile app GGUF iOSAndroid llama.cpp App Store Google Play
ChatterUI Mobile app GGUF Android llama.cpp Download the APK
LLM Farm Mobile app GGUF iOS llama.cpp Not on the App Store or TestFlight for now (checked 2026-11-09).

Which apps can open a model depends on the files published for it: GGUF for llama.cpp-based apps, MLX for Apple’s engine, the original weights for vLLM and SGLang. Compare every app.

Memory needed

Weights plus KV cache and runtime overhead. Leave a little extra free for the operating system and other apps. Try another context length.

Quantization Quality Weights Total at 8K context Total at 32K context Download
Q2_K Low 268 GB 270 GB 271 GB GGUF (Q2_K_L) · folder (opens in a new tab)
Q3_K_M Low 329 GB 330 GB 332 GB GGUF · folder (opens in a new tab)
Q4_K_M Good 409 GB 410 GB 412 GB GGUF · folder (opens in a new tab)
Q5_K_M High 476 GB 478 GB 479 GB GGUF · folder (opens in a new tab)
Q6_K High 550 GB 551 GB 553 GB GGUF · folder (opens in a new tab)
Q8_0 Very high 711 GB 712 GB 714 GB GGUF · folder (opens in a new tab)
F16 Full 1,342 GB 1,343 GB 1,345 GB GGUF · folder (opens in a new tab) Original weights (opens in a new tab)

Best hardware for DeepSeek V3 (671B)

Quality Hardware Quant Memory Speed Runs
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. Apple Mac Studio (2025) (M4 Max or M3 Ultra)Apple M3 Ultra (512 GB) Q3_K_M 361 GB / 384 GB 27.5 tok/s Runs welltight fit
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. Apple Mac Studio (M5 Max or M5 Ultra)Apple M5 Ultra (512 GB) Q3_K_M 361 GB / 384 GB 38.8 tok/s Runs welltight fit
52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. AMD Instinct MI350XAMD Instinct MI350X Q2_K 284 GB / 288 GB 173 tok/s Runs welltight fit
71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. QNAP QAI-h1290FXAMD EPYC 7002 / 7003 series NAS (512 GB) Q4_K_M 448 GB / 471 GB 3.7 tok/s Slowtight fit
52Fair quality (C) · 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. QNAP TDS-h6889AXU-RPIntel Xeon Silver (4100 to 4300 series) NAS (512 GB) Q2_K 307 GB / 471 GB 3.0 tok/s Slow
71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. QNAP TS-2888XIntel Xeon W-2100 series NAS (512 GB) Q4_K_M 448 GB / 471 GB 1.8 tok/s Very slowtight fit
71Great quality (A) · 71 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. TerraMaster U8-725Intel Xeon D-2700 series NAS (512 GB) Q4_K_M 448 GB / 471 GB 1.4 tok/s Very slowtight fit

Check it on your own hardware

Run it on a cluster

Devices that can’t run DeepSeek V3 (671B) alone can run it together. These example clusters do:

Try it with your own devices

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.

Was this page helpful?