Skip to content

DeepSeek V4.1 Flash on 4× H200 NVL: VRAM, speed and fit

DeepSeek V4.1 Flash at FP4 + FP8 runs on the 4× H200 NVL with 1M of context at about 135 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 4× H200 NVL · FP4 + FP8 · Ollama / LM Studio · Chat

GPU memory
481 GB
of 564 GB
Context
1M
auto
Speed
135 tok/s
Very fast
First reply
0.2 s
2K prompt

Where the memory goes

Model
475 GB
Context
2.6 GB
Overhead
2.8 GB
OS
0.0 GB
Free
83.3 GB
Memory
481 GB of 564 GB
Context
1M tokens · auto
Speed
135 tok/s · very fast
First reply
0.2 s for a 2K prompt
Running cost
$33/mo at home · $1,162/mo on a cloud 4× MI300X

Validation: Validated

What DeepSeek V4.1 Flash on 4× H200 NVL was checked against
CheckTargetResultStatus
Fits in GPU memory564 GB usable481 GB neededPass
Headroom5% free83.3 GB free (15%)Pass
Context for chat16K or more1MPass
Writing speed15 tok/s or more135 tok/sPass
First replyunder 3 s, 2K prompt0.2 sPass

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
DeepSeek V4.1 Flash
Parameters
763B, 16B active
Precision
FP4 + FP8
Context
1M, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
4× H200 NVL
GPU can use
564 GB of 564 GB

Running cost

At home
$33/mo
1520 W at the wall
Renting instead
$1,162/mo
4× MI300X at $9.56/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 21, 2026.

How to run it

llama.cpp
llama-server -hf unsloth/DeepSeek-V4.1-Flash-1017-GGUF
vLLM
vllm serve deepseek-ai/DeepSeek-V4.1-Flash
MLX (Mac)
mlx_lm.chat --model mlx-community/DeepSeek-V4.1-Flash-MLX-4bit

DeepSeek V4.1 Flash on similar machines

All machines: DeepSeek V4.1 Flash VRAM requirements. Everything that runs here: what can the 4× H200 NVL run?

Other models that run on 4× H200 NVL

Questions

Can DeepSeek V4.1 Flash run on 4× H200 NVL?

DeepSeek V4.1 Flash at FP4 + FP8 runs on the 4× H200 NVL with 1M of context at about 135 tokens a second.

Report VR-0WB5FH1, calculated Nov 22, 2026 from published specs, not benchmarked.