# DeepSeek-V4-Pro GPU requirements

DeepSeek-V4-Pro has 1.6T parameters. To serve it, plan for about 3,837 GB of GPU memory in 16-bit, 1,919 GB in 8-bit or 959 GB in 4-bit, including about 20% for the KV cache and activations.

It is a mixture of experts: 384 experts, 6 of them active for each token. All of them stay in GPU memory, so plan memory for all 1.6T parameters.

[Deploy it on Arrakis](/console/#/models?model=deepseek-ai/DeepSeek-V4-Pro).

## What it takes to serve it

| Precision | Memory and the smallest setup |
|---|---|
| 16-bit | 3,837 GB: more than one 8-GPU machine |
| 8-bit | 1,919 GB: 8 × [B300](/gpus/b300.html.md), from $36.80 per hour for all 8 |
| 4-bit | 959 GB: 4 × [B300](/gpus/b300.html.md), from $18.40 per hour for all 4 |

## What it takes to fine-tune it

| Method | Memory and the smallest setup |
|---|---|
| QLoRA | 1,119 GB: 8 × [H200](/gpus/h200.html.md), from $32.24 per hour for all 8 |
| LoRA | 3,837 GB: more than one 8-GPU machine |
| Full fine-tuning | 25,581 GB: more than one 8-GPU machine |

## About the model

| | |
|---|---|
| Parameters | `1.6T` |
| Experts | `384` |
| Experts for each token | `6` |
| Context window (tokens) | `1,048,576` |
| Published in | `FP8` |
| License | `mit` |
| Released | `2026-04-22` |
| Weights | [deepseek-ai/DeepSeek-V4-Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro) |

## Questions

**How much GPU memory does DeepSeek-V4-Pro need?** To serve it, about 3,837 GB in 16-bit, 1,919 GB in 8-bit, 959 GB in 4-bit, including about 20% for the KV cache and activations. To fine-tune it, about 1,119 GB with QLoRA, 3,837 GB with LoRA and 25,581 GB for full fine-tuning with Adam, before activations.

**Can I run DeepSeek-V4-Pro on one GPU?** No. Even in 4-bit it needs about 959 GB: 4 B300 GPUs.

**What does it cost to run DeepSeek-V4-Pro?** In 16-bit it needs more than one 8-GPU machine; in 4-bit, 4 B300 GPUs. Each setup on this page shows the lowest price per GPU-hour on Arrakis right now, and you pay by the hour from a prepaid balance.

**How do I deploy DeepSeek-V4-Pro on Arrakis?** Open Models in the Arrakis console, paste deepseek-ai/DeepSeek-V4-Pro, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine.


More models: [/models](/models.html.md).