NVIDIA · Hopper

Best AI Models for NVIDIA H100 SXM (80 GB)

Runs everything up to 120B MoE well.

80 GB HBM33352 GB/s989 TFLOPS FP16700 WHopper 2022

What LLMs can the H100 SXM run?

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 107 t/s	107 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 344 t/s	344 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 284 t/s	284 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 107 t/s	107 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 283 t/s	283 tok/s	Q4_K_M · 22.0 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 344 t/s	344 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 284 t/s	284 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 107 t/s	107 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.5 27BAlibaba Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 344 t/s	344 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 284 t/s	284 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 · 125 t/s	125 tok/s		Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 125 t/s	125 tok/s		Q4_K_M · 17.9 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 283 t/s	283 tok/s		Q4_K_M · 22.0 GB
4	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 107 t/s	107 tok/s	overall	Q4_K_M · 21.7 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 284 t/s	284 tok/s		Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · VRAM
1	Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 668 t/s	668 tok/s	Q4_K_M · 3.4 GB
2	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 399 t/s	399 tok/s	Q4_K_M · 13.1 GB
3	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 344 t/s	344 tok/s	Q4_K_M · 16.5 GB
4	Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 343 t/s	343 tok/s	Q4_K_M · 6.6 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 284 t/s	284 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 107 t/s	107 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 344 t/s	344 tok/s	Q4_K_M · 16.5 GB
5	JEV 27B VLautotrust Q4_K_M · 17.9 GB · 125 t/s	125 tok/s	Q4_K_M · 17.9 GB

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

About the H100 SXM

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

GenerationHopperreleased 2022

H100 SXM strengths and weaknesses

Strengths

Runs 120B MoE models: GPT OSS 120B at 280 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 511 ms

Weaknesses

700 W under load: Plan a 1,400 W PSU; ≈$153 a year at 4 h a day

H100 SXM tokens per second by model size

Output speed depends on memory bandwidth (3352 GB/s).

9BQwen3.5 9B · Q4_K_M · 6.6 GB
343tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
229tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
125tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
283tok/s
70BLlama 3.3 70B Instruct · Q4_K_M · 51.3 GB
46.8tok/s
120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB
280tok/s

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

All 1258 models on the H100 SXM

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 SXM — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
Gemma 4 12B IT Heretic12.0BChat Q4_K_M·264.7 t/s	Q4_K_M	9.0 GB11%	264.7 t/s	131K	S100, Runs great
Ornith 1.0 35B35.1BChatFunctions Q4_K_M·303.6 t/s	Q4_K_M	21.5 GB27%	303.6 t/s	262K	S100, Runs great
Distil Lfm25 Shellper354MChatFunctions Q4_K_M·3756.6 t/s	Q4_K_M	0.8 GB1%	3756.6 t/s	128K	S100, Runs great
Gemma 4 12B IT AEON Abliterated K4 BF1612.0BChatReasoningFunctions Q4_K_M·264.7 t/s	Q4_K_M	9.0 GB11%	264.7 t/s	262K	S100, Runs great
Qwen3 30B A3B Instruct 250730.5BChatFunctions Q4_K_M·279.1 t/s	Q4_K_M	19.4 GB24%	279.1 t/s	262K	S100, Runs great
Moondream3 Preview9.3BVision BF16·106.9 t/s	BF16	23.1 GB29%	106.9 t/s	—	S92, Runs great
Qwen3.8 27B MTPLX Optimized Speed27.4BChat BF16·39.3 t/s	BF16	55.9 GB70%	39.3 t/s	262K	A75, Runs well
VibeThinker 3B3.1BChatMathCodeReasoning Q4_K_M·977.0 t/s	Q4_K_M	2.8 GB3%	977.0 t/s	131K	S100, Runs great
LFM2.5 230M230MChatFunctions Q4_K_M·4357.6 t/s	Q4_K_M	0.6 GB1%	4357.6 t/s	128K	S100, Runs great
Tiny Phi3ForCausalLM 3.5514984Chat Q4_K_M·7262.7 t/s	Q4_K_M	0.5 GB1%	7262.7 t/s	131K	S100, Runs great
SmolVLM2 500M Video Instruct507MVision Q4_K_M·3157.7 t/s	Q4_K_M	0.9 GB1%	3157.7 t/s	8K	S100, Runs great
Nandi Mini 150M153MChat BF16·3458.4 t/s	BF16	0.6 GB1%	3458.4 t/s	2K	S100, Runs great
LFM2.5 8B A1B8.5BChatFunctions Q4_K_M·549.2 t/s	Q4_K_M	5.6 GB7%	549.2 t/s	128K	S100, Runs great
Qwen3.6 27B PRISM PRO DQ27BChat Q4_K_M·122.3 t/s	Q4_K_M	21.4 GB27%	122.3 t/s	—	S95, Runs great
Qwen2.5 7B Pyuigpt7.6BChatFunctions Q4_K_M·436.6 t/s	Q4_K_M	5.8 GB7%	436.6 t/s	33K	S100, Runs great

Rent NVIDIA H100 SXM in the cloud

Hourly rate on demand for one H100 SXM.

Vast.aiMarketplaceLowest price
from$2.20/hr
Deploy
RunPod$5+ bonus credit
from$2.69/hr
Deploy

Cost per million tokens

Output-token cost on an H100 SXM rented at $2.20/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	284 tok/s	$2.16$0.012 input
GLM 4.7 Flash	Q4_K_M	255 tok/s	$2.40$0.013 input
Qwen3.8 27B	Q4_K_M	125 tok/s	$4.89$0.076 input
Muse Glimmer 30B	Q4_K_M	119 tok/s	$5.15$0.082 input
Gemma 4 31B IT	Q4_K_M	107 tok/s	$5.72$0.086 input

H100 SXM specs, power and PSU

Specs

Architecture: Hopper
VRAM: 80 GB HBM3
Memory bandwidth: 3352.0 GB/s
FP16 (dense): 989.40 TFLOPS
CUDA cores: 16,896
Tensor cores: 528
Compute capability: 9.0
TDP: 700 W
Released: 2022-09-01

H100 SXM FAQ

What is the largest model the H100 SXM 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 29.9 tok/s.

Can the H100 SXM 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 SXM give?

About 125 tok/s on Qwen3.8 27B at Q4_K_M, with the first token after 0.5 s on a 4K-token prompt.

What power supply does the H100 SXM need?

Plan a 1400 W PSU (2× the 700 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 125 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 RTX 5090