NVIDIA · Hopper

Best AI Models for NVIDIA H200 SXM (141 GB)

Runs everything up to 120B MoE well.

141 GB HBM3e4800 GB/s990 TFLOPS FP16700 WHopper 2024

What LLMs can the H200 SXM run?

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 153 t/s	153 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 395 t/s	395 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 301 t/s	301 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 153 t/s	153 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 320 t/s	320 tok/s	Q4_K_M · 22.0 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 395 t/s	395 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 301 t/s	301 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 153 t/s	153 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.5 27BAlibaba Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 395 t/s	395 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 301 t/s	301 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 · 179 t/s	179 tok/s		Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 179 t/s	179 tok/s		Q4_K_M · 17.9 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB · 320 t/s	320 tok/s		Q4_K_M · 22.0 GB
4	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 153 t/s	153 tok/s	overall	Q4_K_M · 21.7 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 301 t/s	301 tok/s		Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · VRAM
1	Qwen3.5 4BAlibabaHow to run → Q4_K_M · 3.4 GB · 957 t/s	957 tok/s	Q4_K_M · 3.4 GB
2	Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB · 491 t/s	491 tok/s	Q4_K_M · 6.6 GB
3	GPT OSS 20BOpenAI Q4_K_M · 13.1 GB · 465 t/s	465 tok/s	Q4_K_M · 13.1 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 395 t/s	395 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB · 301 t/s	301 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · VRAM
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB · 179 t/s	179 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB · 153 t/s	153 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB · 395 t/s	395 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Super VL 120B A12B BF16NVIDIA Q4_K_M · 87.7 GB · 193 t/s	193 tok/s	Q4_K_M · 87.7 GB

See all 404 models on the H200 SXM →Best quant that fits · speeds estimated on this card

About the H200 SXM

The H200 SXM runs everything up to 120B MoE well. Hopper generation, released 2024.

GenerationHopperreleased 2024

H200 SXM strengths and weaknesses

Strengths

Runs 120B MoE models: GPT OSS 120B at 323 tok/s in 70.8 GB
Room for long context: Llama 3.3 70B Instruct at Q4 uses 51.3 GB, leaving 89.7 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

H200 SXM tokens per second by model size

Output speed depends on memory bandwidth (4800 GB/s).

9BQwen3.5 9B · Q4_K_M · 6.6 GB
491tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
328tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
179tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
320tok/s
70BLlama 3.3 70B Instruct · Q4_K_M · 51.3 GB
67.0tok/s
120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB
323tok/s

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

All 1295 models on the H200 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 H200 SXM — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
MN 12B Mag Mell R112.2BChatReasoning Q4_K_M·386.6 t/s	Q4_K_M	11.0 GB8%	386.6 t/s	1024K	S100, Runs great
LLaMmlein 1B Prerelease1.1BChat Q4_K_M·3089.1 t/s	Q4_K_M	1.0 GB1%	3089.1 t/s	2K	S100, Runs great
Falcon 11B11.1BChat Q4_K_M·425.6 t/s	Q4_K_M	8.5 GB6%	425.6 t/s	8K	S100, Runs great
Qwen2.5 3B Instruct3.1BChatFunctions Q4_K_M·1399.1 t/s	Q4_K_M	2.8 GB2%	1399.1 t/s	33K	S100, Runs great
Qwen2.5 1.5B Instruct1.5BChatFunctions Q4_K_M·2437.5 t/s	Q4_K_M	1.7 GB1%	2437.5 t/s	33K	S100, Runs great
EuroLLM 1.7B Instruct1.7BChat Q4_K_M·2080.0 t/s	Q4_K_M	1.7 GB1%	2080.0 t/s	4K	S100, Runs great
Shieldgemma 9B9.2BChat Q4_K_M·511.5 t/s	Q4_K_M	8.8 GB6%	511.5 t/s	—	S100, Runs great
Falcon Mamba Tiny Dev9MChat BF16·156000.0 t/s	BF16	0.4 GB0%	156000.0 t/s	—	S100, Runs great
ERNIE 4.5 0.3B Paddle361MChat BF16·3000.0 t/s	BF16	1.2 GB1%	3000.0 t/s	131K	S100, Runs great
Qwen2.5 Omni 3B MNN3BChat BF16·472.7 t/s	BF16	8.6 GB6%	472.7 t/s	—	S100, Runs great
Gemma 3 12B IT12.2BVision Q4_K_M·388.1 t/s	Q4_K_M	10.9 GB8%	388.1 t/s	33K	S100, Runs great
Llama 3.3 70B Instruct Abliterated70.6BChatFunctions Q4_K_M·72.1 t/s	Q4_K_M	48 GB34%	72.1 t/s	131K	S86, Runs great
Ministral 3 3B Reasoning 25123.8BReasoningFunctions Q4_K_M·1126.4 t/s	Q4_K_M	3.9 GB3%	1126.4 t/s	262K	S100, Runs great
Llama 3.2 1B Instruct1.2BChatFunctions Q4_K_M·3804.9 t/s	Q4_K_M	2.3 GB2%	3804.9 t/s	131K	S100, Runs great
Qwen2.5 1.5B Quantized.w8a81.8BChatFunctions Q4_K_M·2181.8 t/s	Q4_K_M	1.8 GB1%	2181.8 t/s	33K	S100, Runs great

Rent NVIDIA H200 SXM in the cloud

Hourly rate on demand for one H200 SXM.

RunPod$5+ bonus creditLowest price
from$3.59/hr
Deploy
Vast.aiMarketplace
from$4.80/hr
Deploy

Cost per million tokens

Output-token cost on an H200 SXM rented at $3.59/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	301 tok/s	$3.31$0.020 input
Qwen3.8 27B	Q4_K_M	179 tok/s	$5.57$0.12 input
Muse Glimmer 30B	Q4_K_M	170 tok/s	$5.86$0.13 input
Gemma 4 31B IT	Q4_K_M	153 tok/s	$6.52$0.14 input
Qwen3.8 Flash Next	Q4_K_M	69 tok/s	$14.47$0.28 input

H200 SXM specs, power and PSU

Specs

Architecture: Hopper
VRAM: 141 GB HBM3e
Memory bandwidth: 4800.0 GB/s
FP16 (dense): 989.50 TFLOPS
CUDA cores: 16,896
Tensor cores: 528
Compute capability: 9.0
TDP: 700 W
Released: 2024-03-01

H200 SXM FAQ

What is the largest model the H200 SXM can run?

Qwen3 235B A22B Thinking 2507 (235.1B) is the largest at 4-bit or better: at Q4_K_S it uses 134.1 GB of 141 GB, and runs at about 109 tok/s. At 3-bit, a bigger model fits too: DeepSeek V4 Flash (290.9B) at IQ3_M, 131.4 GB.

Can the H200 SXM run Qwen3.8 27B?

Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of 141 GB, and it is the best model that runs well on this card.

How many tokens per second does the H200 SXM give?

About 179 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 H200 SXM need?

Plan a 1400 W PSU (2× the 700 W TDP).

Is 141 GB enough for coding and agents?

Yes for coding: Qwen3.8 27B, the best of the 14 coding models that fit, runs at about 179 tok/s.

Best coding LLMs 48 GB VRAM GPUs with 141 GB or more RTX 5090 RTX 4090 RTX 3090