DeepSeek-V3

671B MoE

DeepSeek

DeepSeek-V3 frontier MoE (~37B active / 671B total). MLA + FP8; multi-node GPU cluster required at Q4.

1.1M HF downloads4215 likesdeepseek-ai/DeepSeek-V3· stats from 9/23/2026
Pro GPU

164K

Max Context

2

Quant Variants

GGUF Q4_K_M

Best Quality

98.0%

Accuracy Retained

Quantization Variants

Per-quant VRAM, quality loss, and inference speed on RTX 4090

Measured = site benchmarks · Estimated = formula · Community = public reports

FormatLevelBPWVRAMPPL LossSpeedSourceActions
GGUFQ4_K_M4.85385.0 GB2.0%4 tok/sEstimated
CalcHF
GGUFQ3_K_M3.87310.0 GB4.5%5 tok/sEstimated
CalcHF

Running DeepSeek-V3 locally

At Q4_K_M and 4K of context, DeepSeek-V3 needs about 426.1 GB — 386.4 GB of weights, 0.95 GB of KV cache and a 38.7 GB activation buffer. No card in this index clears that comfortably, so it is a multi-GPU or CPU-offload proposition. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs.

What longer context costs

Going from 4K to 32K adds about 6.68 GB, taking the total to 433.5 GB. Weights do not move with context — only the KV cache does, and it grows linearly, so this is the number to watch when planning for long documents. The model's native window is 160K; holding all of it at Q4_K_M would need about 467 GB.

Which build to download

This index only tracks DeepSeek-V3 in GGUF, across 2 levels (Q4_K_M, Q3_K_M). Q4_K_M carries the lowest published perplexity loss at 2.0%. The fastest level measured here is Q3_K_M at 5 tok/s on an RTX 4090, batch 1. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 and CUDA only.

Common questions

How much VRAM does DeepSeek-V3 need?
About 426.1 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 433.5 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.
Can DeepSeek-V3 run on a single GPU?
Not comfortably on any single card in this index at Q4_K_M. It needs about 426.1 GB, which means splitting across GPUs or offloading layers to system RAM.
Which quantization of DeepSeek-V3 should I use?
Q4_K_M has the lowest published quality loss (2.0%), and Q4_K_M is the level most people run. All 2 levels in the index are Q4_K_M, Q3_K_M.

Where this model fits

Sized at Q4_K_M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.

Ships in
GGUF

This page's figures change when the model or the runtime does.Last updated 2026-09-23 RSS → /feed.xml