RunLocalAI
nvidia
GPU
141GB VRAM
workstation

NVIDIA H200

Hopper refresh — 141GB HBM3e at ~4.8 TB/s. Datacenter-class; rentable on RunPod, Lambda, etc.

Released 2024

Overview

Hopper refresh — 141GB HBM3e at ~4.8 TB/s. Datacenter-class; rentable on RunPod, Lambda, etc.

Specs

VRAM141 GB
Power draw700 W
Released2024
MSRP$31000
Backends
CUDA

Models that fit

Open-weight models small enough to run on NVIDIA H200 with usable context.

Llama 3.1 8B Instruct
8B · llama
Qwen 3 30B-A3B
30B · qwen
Qwen 2.5 Coder 32B Instruct
32B · qwen
Llama 3.3 70B Instruct
70B · llama
Qwen 3 32B
32B · qwen
Gemma 4 31B Dense
31B · gemma
Qwen 3 8B
8B · qwen
DeepSeek R1 Distill Llama 70B
70B · deepseek

Frequently asked

What models can NVIDIA H200 run?

With 141GB VRAM, the NVIDIA H200 runs 70B models in 4-bit quantization, plus everything smaller. See the model list below for tested combinations.

Does NVIDIA H200 support CUDA?

Yes — NVIDIA H200 is an NVIDIA card with full CUDA support, the most mature local-AI backend. llama.cpp, Ollama, vLLM, and ExLlamaV2 all run natively.

Reviewed by RunLocalAI Editorial. See our editorial policy for how we research and verify hardware specifications.