NVIDIA · Hopper

Best AI Models for NVIDIA H100 PCIe (80 GB)

Runs everything up to 120B MoE well.

80 GB HBM2e2039 GB/s757 TFLOPS FP16350 WHopper 2022

What LLMs can the H100 PCIe run?

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 65.0 t/s	65.0 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 269 t/s	269 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 253 t/s	253 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 65.0 t/s	65.0 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 228 t/s	228 tok/s	Q4_K_M · 22.0 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 269 t/s	269 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 253 t/s	253 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 65.0 t/s	65.0 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.5 27BAlibaba Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 269 t/s	269 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 253 t/s	253 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Math Score	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s		Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s		Q4_K_M · 17.9 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 228 t/s	228 tok/s		Q4_K_M · 22.0 GB
4	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 65.0 t/s	65.0 tok/s	overall	Q4_K_M · 21.7 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 253 t/s	253 tok/s		Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · VRAM
1	Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 407 t/s	407 tok/s	Q4_K_M · 3.4 GB
2	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 305 t/s	305 tok/s	Q4_K_M · 13.1 GB
3	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 269 t/s	269 tok/s	Q4_K_M · 16.5 GB
4	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 253 t/s	253 tok/s	Q4_K_M · 19.3 GB
5	Qwen3 30B A3B Thinking 2507Alibaba Q4_K_M · 19.4 GB · 239 t/s	239 tok/s	Q4_K_M · 19.4 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 65.0 t/s	65.0 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 269 t/s	269 tok/s	Q4_K_M · 16.5 GB
5	JEV 27B VLautotrust Q4_K_M · 17.9 GB · 76.1 t/s	76.1 tok/s	Q4_K_M · 17.9 GB

See all 383 models on the H100 PCIe →Best quant that fits · speeds estimated on this card

About the H100 PCIe

The H100 PCIe runs everything up to 120B MoE well. Hopper generation, released 2022.

GenerationHopperreleased 2022

H100 PCIe strengths and weaknesses

Strengths

Runs 120B MoE models: GPT OSS 120B at 218 tok/s in 70.8 GB
Room for long context: Llama 3.3 70B Instruct at Q4 uses 51.3 GB, leaving 28.7 GB free
Reads long prompts fast: A 4K-token prompt on Qwen3.8 27B starts answering in 669 ms

H100 PCIe tokens per second by model size

Output speed depends on memory bandwidth (2039 GB/s).

9BQwen3.5 9B · Q4_K_M · 6.6 GB
208tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
140tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
76.1tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
228tok/s
70BLlama 3.3 70B Instruct · Q4_K_M · 51.3 GB
28.5tok/s
120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB
218tok/s

Estimated speeds: Q4_K_M, 4K context, one user. How we estimate

All 383 models on the H100 PCIe

Featured only

Best quant that fits, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with NVIDIA H100 PCIe — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
Gemma 4 E4B IT8.0BChatFunctions Q4_K_M·249.1 t/s	Q4_K_M	5.6 GB7%	249.1 t/s	131K	S100, Runs great
LFM3 1.7B1.7BChatFunctions Q4_K_M·920.4 t/s	Q4_K_M	2.1 GB3%	920.4 t/s	262K	S100, Runs great
MiniCPM V 4.7 1B1.6BVisionFunctions Q4_K_M·1035.4 t/s	Q4_K_M	1.6 GB2%	1035.4 t/s	262K	S100, Runs great
LFM3 Omni 2B2.2BVisionChatFunctions Q4_K_M·779.6 t/s	Q4_K_M	2.4 GB3%	779.6 t/s	128K	S100, Runs great
Ornith 1.5 122B125.1BChatFunctions Q4_K_M·118.5 t/s	Q4_K_M	75.5 GB94%	118.5 t/s	262K	B62, Decent
Kolibri 1 BF1678.1BChatReasoningFunctions Q4_K_M·231.0 t/s	Q4_K_M	47.9 GB60%	231.0 t/s	262K	S100, Runs great
ZAYA1 VL 8B9.7BVision Q4_K_M·393.4 t/s	Q4_K_M	7.5 GB9%	393.4 t/s	33K	S100, Runs great
GPT OSS 20B20.9BChatFunctions Q4_K_M·305.4 t/s	Q4_K_M	13.1 GB16%	305.4 t/s	131K	S100, Runs great
LFM3 VL 2B1.9BVisionChatFunctions Q4_K_M·855.1 t/s	Q4_K_M	2.3 GB3%	855.1 t/s	128K	S100, Runs great
Clef Flash9.4BVisionFunctions Q4_K_M·213.4 t/s	Q4_K_M	6.5 GB8%	213.4 t/s	262K	S100, Runs great
Holo4 35B A3B35.1BVisionFunctions Q4_K_M·250.1 t/s	Q4_K_M	21.5 GB27%	250.1 t/s	262K	S100, Runs great
InternVL3 5 GPT OSS 20B A4B Preview HF21.2BVision Q4_K_M·292.3 t/s	Q4_K_M	13.6 GB17%	292.3 t/s	131K	S100, Runs great
Ling Flash 2.0102.9BChatFunctions Q4_K_M·219.1 t/s	Q4_K_M	63.1 GB79%	219.1 t/s	33K	S100, Runs great
Magistral Small 250924.0BChat Q4_K_M·87.6 t/s	Q4_K_M	18.1 GB23%	87.6 t/s	131K	S89, Runs great
North Mini Code 1.030.5BChatCodeFunctions Q4_K_M·236.7 t/s	Q4_K_M	19.4 GB24%	236.7 t/s	500K	S100, Runs great

Rent NVIDIA H100 PCIe in the cloud

Hourly rate on demand for one H100 PCIe.

Vast.aiMarketplace
from$2.76/hr
Deploy

Cost per million tokens

Output-token cost on an H100 PCIe rented at $2.76/hr, one request at a time. Batch several users on the same card and the cost per token drops several-fold.

Model	Quant	Speed	Per 1M output
NVIDIA Nemotron 3.5 Lightning 30B A3B BF16	Q4_K_M	253 tok/s	$3.02$0.015 input
GLM 4.7 Flash	Q4_K_M	209 tok/s	$3.67$0.021 input
Qwen3.8 27B	Q4_K_M	76 tok/s	$10.06$0.12 input
Muse Glimmer 30B	Q4_K_M	72 tok/s	$10.59$0.13 input
Gemma 4 31B IT	Q4_K_M	65 tok/s	$11.78$0.14 input

H100 PCIe specs, power and PSU

Specs

Architecture: Hopper
VRAM: 80 GB HBM2e
Memory bandwidth: 2039.0 GB/s
FP16 (dense): 756.50 TFLOPS
CUDA cores: 14,592
Tensor cores: 456
Compute capability: 9.0
TDP: 350 W
Released: 2022-09-01

H100 PCIe FAQ

What is the largest model the H100 PCIe can run?

Mistral Medium 3.5 128B (127.7B) is the largest at 4-bit or better: at Q4_K_S it uses 78.0 GB of 80 GB, and runs at about 18.2 tok/s.

Can the H100 PCIe run Qwen3.8 27B?

Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of 80 GB, and it is the best model that runs well on this card.

How many tokens per second does the H100 PCIe give?

About 76.1 tok/s on Qwen3.8 27B at Q4_K_M, with the first token after 0.7 s on a 4K-token prompt.

What power supply does the H100 PCIe need?

Plan a 700 W PSU (2× the 350 W TDP).

Is 80 GB enough for coding and agents?

Yes for coding: Qwen3.8 27B, the best of the 12 coding models that fit, runs at about 76 tok/s. For agents, not yet: no agent-scored model fits in less than 96 GB.

Best coding LLMs 48 GB VRAM GPUs with 80 GB or more