H200 141GB SXM5 vs V100 16GB PCIe for Local LLMs

The H200 141GB SXM5 has about 433% more memory bandwidth (4,800 vs 900 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 433% faster on the H200 141GB SXM5. The H200 141GB SXM5 fits more of the catalog at Q4/8K — 59 vs 25 of 73 models (141 vs 16 GB VRAM). Largest model that fits the H200 141GB SXM5 but not the V100 16GB PCIe: Qwen3-235B-A22B-Instruct-2507 (235.1B). MSRP is $35,000 for the H200 141GB SXM5 vs $8,000 for the V100 16GB PCIe. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. Not retail-shoppable: H200 141GB SXM5, V100 16GB PCIe (datacenter-class; cloud rental is the realistic way to use it). In the Q4/8K head-to-head below (49 shared models), the H200 141GB SXM5 posts the higher tok/s in 49 rows vs 0. Methodology.

Spec comparison

	H200 141GB SXM5	V100 16GB PCIe
Vendor	NVIDIA	NVIDIA
VRAM	141 GB	16 GB
Memory type	HBM3e	HBM2
Memory bandwidth	4,800 GB/s	900 GB/s
TDP	700 W	250 W
MSRP (list)	$35,000	$8,000
Buy	—	—

MSRP is the launch list price, not live retail. Buy links are Amazon affiliate links — the price after you click is the live price.

Q4 / 8K head-to-head — top 15 of 49 shared models

Model	Params (B)	H200 141GB SXM5 tok/s	H200 141GB SXM5 fits	V100 16GB PCIe tok/s	V100 16GB PCIe fits	Winner
MiniCPM5-1B	1.08	4823.4 est.	Yes	895.6 est.	Yes	H200 141GB SXM5
Qwen3.5-2B	2.0	2604.7 est.	Yes	483.6 est.	Yes	H200 141GB SXM5
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	2264.9 est.	Yes	420.5 est.	Yes	H200 141GB SXM5
Qwen3.6-35B-A3B	36.0	1736.4 est.	Yes	316.9 est.	No	H200 141GB SXM5
Qwen3.5-35B-A3B	34.7	1736.4 est.	Yes	339.8 est.	No	H200 141GB SXM5
GLM-4.7-Flash	31.2	1736.4 est.	Yes	313.0 est.	No	H200 141GB SXM5
Qwen3-30B-A3B-Instruct-2507	30.5	1736.4 est.	Yes	405.4 est.	No	H200 141GB SXM5
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	1736.4 est.	Yes	434.0 est.	No	H200 141GB SXM5
Qwen3-Coder-Next	79.7	1736.4 est.	Yes	67.0 est.	No	H200 141GB SXM5
Ornith-1.0-35B	35.0	1736.4 est.	Yes	334.2 est.	No	H200 141GB SXM5
Qwen3-Next-80B-A3B-Instruct	81.3	1736.4 est.	Yes	64.5 est.	No	H200 141GB SXM5
qwen35b-a3b-fable-sft-abliterated	34.7	1736.4 est.	Yes	325.2 est.	No	H200 141GB SXM5
Llama-3.2-3B-Instruct	3.2	1627.9 est.	Yes	302.3 est.	Yes	H200 141GB SXM5
gpt-oss-20b	21.5	1447.0 est.	Yes	268.7 est.	Yes	H200 141GB SXM5
gemma-4-26B-A4B-it	25.2	1370.9 est.	Yes	254.5 est.	Yes	H200 141GB SXM5

est. = bandwidth-based estimate; measured = curated real benchmark that overrides the estimate. Fits = weights + KV(8K) + 1 GB overhead ≤ 95% of VRAM.

Fit coverage at Q4 / 8K

	H200 141GB SXM5	V100 16GB PCIe
Catalog models that fit	59 of 73	25 of 73
Largest exclusive fit	Qwen3-235B-A22B-Instruct-2507 (235.1B)	—

H200 141GB SXM5 details → · V100 16GB PCIe details →