NVIDIA8x H100 SXM5Server

Best AI Models for NVIDIA DGX H100

Memory:640 GB Unified·Bandwidth:26800.0 GB/s·GPU Cores:135168 GPU cores·CPU Cores:112 CPU cores

640 GB unified − 3.5 GB OS overhead = 636.5 GB available for AI models

With 640 GB of memory, this is a high-end configuration for local AI. You can comfortably run most open-source LLMs including large 70B parameter models at good quantization levels, making it one of the best setups for serious local AI work.

At this memory tier, nearly every popular open-source model is within reach. You can run Llama 3 70B at Q4_K_M or even Q5_K_M quantization with room to spare, handle coding assistants like DeepSeek Coder 33B at high quality, and easily run any 7B–30B model at full or near-full precision. Context windows remain generous even with larger models, so multi-turn conversations and long-document processing work smoothly.

Runs Well

  • 70B models (Llama 3 70B, Qwen 72B) at Q4–Q5
  • 30B models at Q6–Q8 quality
  • 7B–14B models at full FP16 precision
  • Vision models (LLaVA, CogVLM) without compromise

Challenging

  • Mixture-of-experts models like Mixtral 8x22B at higher quants
  • 120B+ models still require lower quantizations

What LLMs Can NVIDIA DGX H100 Run?

175 models · 155 excellent · 13 good

Showing compatibility for NVIDIA DGX H100

LLM models compatible with NVIDIA DGX H100 — ranked by performance
ModelVRAMGrade
GLM 4.5358.3B
Q4_K_M·80.8 t/s tok/s·131K ctx·RUNS GREAT
215.6 GBS88
GLM 4.6356.8B
Q4_K_M·81.1 t/s tok/s·203K ctx·RUNS GREAT
214.7 GBS88
BF16·89.4 t/s tok/s·RUNS GREAT
194.9 GBS89
Inkling952.4B
Q4_K_M·30.4 t/s tok/s·DECENT
572.1 GBB59
Qwen3 235B A22B235.1B
Q4_K_M·123.1 t/s tok/s·41K ctx·RUNS GREAT
141.6 GBS95
Hy3298.8B
Q4_K_M·96.8 t/s tok/s·262K ctx·RUNS GREAT
179.9 GBS90
MiniMax M2228.7B
Q4_K_M·126.4 t/s tok/s·197K ctx·RUNS GREAT
137.8 GBS95
GPT OSS 120B120.4B
Q4_K_M·239.8 t/s tok/s·131K ctx·RUNS GREAT
72.7 GBS100
Step 3.7 Flash201.4B
Q4_K_M·131.1 t/s tok/s·262K ctx·RUNS GREAT
132.9 GBS95
Q4_K_M·123.1 t/s tok/s·262K ctx·RUNS GREAT
141.6 GBS95
BF16·105.1 t/s tok/s·262K ctx·RUNS GREAT
165.8 GBS92
Q4_K_M·123.1 t/s tok/s·262K ctx·RUNS GREAT
141.6 GBS95
Q4_K_M·175.1 t/s tok/s·1049K ctx·RUNS GREAT
99.5 GBS99
Q4_K_M·132.4 t/s tok/s·200K ctx·RUNS GREAT
131.6 GBS95
Q4_K_M·233.4 t/s tok/s·262K ctx·RUNS GREAT
74.7 GBS100
DeepSeek v2235.7B
Q4_K_M·120.8 t/s tok/s·164K ctx·RUNS GREAT
144.3 GBS95

NVIDIA DGX H100 Specifications

Brand
NVIDIA
Chip
8x H100 SXM5
Type
Server
Unified Memory
640 GB
Memory Bandwidth
26800.0 GB/s
GPU Cores
135168
CPU Cores
112
TDP
10200 W
Interconnect
NVSwitch 4th-gen, 900 GB/s NVLink per GPU
MSRP
$482,000
Release Date
2022-09-01

Get Started

Ollama (Recommended)

$curl -fsSL https://ollama.com/install.sh | sh
$ollama run llama3:8b

LM Studio

LM Studio

Download LM Studio, search for a model, and run it with one click.

Prompt Processing

Estimated for GLM 5, the compute-bound phase that reads your prompt before the first reply token appears.

2,362.4tok/s prefill

Short chat

217 ms

512 tok prompt

Long chat

1.7 s

4,096 tok prompt

Document / codebase

13.9 s

32,768 tok prompt

Prefill is compute-bound and a different number from the decode tok/s shown elsewhere on this page — how prompt processing works →

Performance figures are estimates calibrated as of 2026-07-30 see calibration basis →

Devices to Consider

Similar devices and upgrades with more memory or higher bandwidth

Frequently Asked Questions

Can NVIDIA DGX H100 run DeepSeek R1?

Yes, the NVIDIA DGX H100 with 640 GB unified memory can run DeepSeek R1, GLM 5.2, DeepSeek V3.2, and 1789 other models. 1746 models achieve excellent performance, and 34 run at good quality. Apple Silicon's unified memory architecture lets the GPU access the full memory pool without copying data, making it efficient for AI workloads.

How much memory is available for AI on NVIDIA DGX H100?

The NVIDIA DGX H100 has 640 GB unified memory. After macOS reserves ~3.5 GB for the operating system, approximately 636.5 GB is available for AI models. Unlike discrete GPUs where VRAM is separate from system RAM, Apple Silicon shares one memory pool between the CPU and GPU — this means no data copying overhead, but you share memory with macOS and open apps.

Is NVIDIA DGX H100 good for AI?

With 640 GB unified memory and 26800.0 GB/s bandwidth, the NVIDIA DGX H100 is excellent for running local AI models. It supports 1780 models at good quality or better. This is a premium configuration — you can run large 30B+ parameter models at good quality, and most 7B models at maximum quality. Ideal for professional AI workloads.

What's the best model for NVIDIA DGX H100?

The top-rated models for the NVIDIA DGX H100 are DeepSeek R1, GLM 5.2, DeepSeek V3.2. With this much memory, you can prioritize quality — use higher quantizations (Q5/Q6) for better output, or run larger 30B+ models for more capable reasoning.

How fast is NVIDIA DGX H100 for AI inference?

With 26800.0 GB/s memory bandwidth, the NVIDIA DGX H100 achieves approximately 4169 tok/s on a 7B model at Q4_K_M — that's very fast, well above conversational speed. A 14B model runs at ~2084 tok/s. Apple Silicon achieves high efficiency (~70%) thanks to unified memory — there's no PCIe bottleneck between CPU and GPU.

tok/s = (26800 GB/s ÷ model GB) × efficiency

Apple Silicon achieves ~70% bandwidth efficiency thanks to unified memory and Metal acceleration.

Estimated speed on NVIDIA DGX H100

~42 tok/s
~38 tok/s
~42 tok/s
~42 tok/s

Real-world results typically within ±20%.

Learn more about tok/s estimation →

Can I run AI offline on NVIDIA DGX H100?

Yes — once you download a model, it runs entirely on the NVIDIA DGX H100 without internet. Applications like Ollama and LM Studio make it straightforward to download, manage, and run models locally. All your conversations stay private on your device with zero data sent to external servers. This is one of the key advantages of local AI: complete privacy, no API costs, and no rate limits.