DeepSeek V4 Pro GPU requirements: 865 GB at FP4+FP8, from $8.72/h

Memory by format

How much VRAM DeepSeek V4 Pro needs, format by format.

Weights are parameters × bytes per parameter, except the native FP4+FP8 row, which is the size of the published checkpoint. The KV cache of this architecture is not estimated (see below), so only the weights are shown.

GPUs by memory

Format	Weights	Cheapest node, 32K
BF16Published weights, fine-tuning	3.2 TB	more than one node
FP8Hopper, Ada, Blackwell, MI300	1.6 TB	8× MI325X · $8.72/h
FP4+FP8 (native)AWQ, GPTQ, GGUF Q4 · the published checkpoint	865 GB	8× B200 SXM · $13.52/h

Arithmetic on published facts, not a benchmark.

The serving engine adds its own buffers, and an 8-bit KV cache halves the context columns. We keep 10% of a node's memory free when we say a model fits. Speed is a separate question: it follows memory bandwidth and, for a mixture of experts, the active parameters.

Where it runs

GPUs that run DeepSeek V4 Pro, cheapest first.

For each format, the smallest node of each GPU model that holds the weights and a 32K context, sorted by spot price. One row per GPU model; larger nodes of the same GPU work too.

All prices

Format	GPU	GPU memory	Spot	On-demand	730 h spot	Action
FP4+FP8 (native)	8× B200 SXMBlackwell · 865 GB needed	1440 GB	$13.52	$19.92	$9,869.60	Deploy
FP8	8× MI325XCDNA 3 · 1.6 TB needed	2048 GB	$8.72	$12.72	$6,365.60	Deploy

Spot capacity can be reclaimed with a 2-minute notice; for a live endpoint, use on-demand. How spot works · serving a language model.

Context and the KV cache

What a long context costs in memory.

Compressed attention (per-layer compression ratios of 4 and 128 plus a 128-token window); its cache is a small fraction of a standard model's and we publish no estimate we cannot check.

No cache estimate for this architecture.

The cache is a small fraction of a standard model's at the same length, but its size depends on details we cannot check from the published configuration, so we do not print a number. The node recommendations on this page are sized on the weights plus the 10% margin.

Serving several users

KV cache

Every concurrent request keeps its own cache, so memory for context grows with users × context length. A server batching 8 requests of 32K needs eight times the single-sequence figure.

Paged attention (vLLM, SGLang) allocates the cache in blocks as requests grow, so a node holds more short requests than the worst case suggests.

Model facts

DeepSeek V4 Pro, as published.

Developer: DeepSeekThe large DeepSeek V4 model, 1M-token context
Parameters: 1.6T total, 49B active per token
Architecture: Mixture of experts, 61 layersCompressed attention (per-layer compression ratios of 4 and 128 plus a 128-token window); its cache is a small fraction of a standard model's and we publish no estimate we cannot check
Context window: 1,048,576 tokensAs set in the published configuration
Published format: FP4+FP8, 865 GB checkpoint
Licence: MIT
Released: 22 April 2026
Source: deepseek-ai/DeepSeek-V4-Proconfig.json read 26 October 2026

Serve it with vLLM

template: vLLM

On 8× MI325X, from the instance terminal:

terminal

$ vllm serve deepseek-ai/DeepSeek-V4-Pro \
    --tensor-parallel-size 8 \
    --max-model-len 32768

Very recent architectures need a recent vLLM; the model card gives the minimum version and any extra flags.

Related models

Other models of a similar size.

All 64 models

DeepSeek V3.2 671B · 37B active · 343 GB at 4-bit $0.48/h

V3 architecture with DeepSeek Sparse Attention

DeepSeek R1 671B · 37B active · 342 GB at 4-bit $0.48/h

Reasoning model, the R1-0528 checkpoint

DeepSeek V4.1 Flash 552B · 510 GB at FP4+FP8 $1.52/h

Multimodal mixture of experts: 552B backbone plus 196B of Engram memory, 8B or 16B activated per token (model card), 1M-token context

DeepSeek V4 Flash 284B · 13B active · 167 GB at FP4+FP8 $0.38/h

Fast DeepSeek V4 mixture of experts with a 1M-token context, the 0731 update

FAQ

DeepSeek V4 Pro on a GPU — the questions that come up

How much VRAM does DeepSeek V4 Pro need?

For the weights alone: 3.2 TB at BF16, 1.6 TB at FP8 and 865 GB at FP4+FP8. Its attention scheme keeps a much smaller cache than a standard model and we do not publish an estimate we cannot check; keep a margin for the cache on top of the weights.

Can DeepSeek V4 Pro run on a single GPU?

One RTX 4090 (24 GB): no. One RTX 5090 (32 GB): no. One H100 SXM (80 GB): no. One H200 SXM (141 GB): no. One B200 SXM (180 GB): no. Those answers assume a 32K context and the widest format that still fits.

What is the cheapest way to run DeepSeek V4 Pro?

Here, 8× MI325X at FP8: $8.72 per hour on spot and $12.72 on-demand, billed per minute — $6,365.60 for a 730-hour month on spot. That node has 2048 GB of GPU memory for the 1.6 TB this setup needs.

Does DeepSeek V4 Pro need less memory because only 49B parameters are active?

No. A mixture-of-experts model routes each token through a few experts, but every expert has to be resident in GPU memory, so the memory bill follows the 1.6T total. What the 49B active parameters change is the work per token: generation reads far less of the weights each step than a dense model of the same size would, which is why it is fast for its size.

Which format should I run it in?

DeepSeek V4 Pro is published in FP4+FP8, a 865 GB checkpoint — the format its developers serve it in, so it is the safe default. BF16 keeps the published weights untouched and is the format for fine-tuning. FP8 halves the memory and runs in hardware on Hopper, Ada, Blackwell and MI300-class GPUs. 4-bit quantisation (AWQ, GPTQ, GGUF) quarters it, at a quality cost that is larger on small models than on big ones; check it against your own evaluation before relying on it.

How do I deploy it?

Pick the node in the table above and launch it with the vLLM template; the model downloads from Hugging Face (deepseek-ai/DeepSeek-V4-Pro) on first start. Spread the weights over the node with --tensor-parallel-size equal to the GPU count, and set --max-model-len to the context you need. New architectures need a recent vLLM: check the model card for the minimum version.

Sizing another model? GPU requirements for 64 open models · serving a language model · fine-tuning.

Get started

DeepSeek V4 Pro on 8× MI325X for $8.72 an hour.

Launch the node with the vLLM template, point it at deepseek-ai/DeepSeek-V4-Pro, stop it when you are done. Prepaid, billed by the minute.

Deploy 8× MI325X All open models

Billed per minute from the moment the instance is reachable. Minimum credit $30, no subscription.