OpenModelsHQ

H100 SXM

NVIDIA · Data center GPU

H100 SXM has 80 GB of memory and 3,350 GB/s of memory bandwidth. 409 open models in our directory fit at 4-bit.

Usable for models, before runtime overhead
80GB
Listed memory, before overhead
Total memory
80 GB
Bandwidth
3,350 GB/s
Models that fit at 4-bit
409
Within 72.0 GB, 90% of usable

NVIDIA lists the power limit as up to 700 W (configurable).

Planning a multi-GPU deployment? Check NVLink domains and network topology as well as memory fit. A GPU specification alone does not establish the interconnect in a server or cloud allocation.

What fits by category
On H100 SXM at 4-bit, 409 models fit, 6 are tight and 76 are too big, across 6 categories. Fits means the estimate is within 90% of usable memory; tight is between 90% and 100%. Models without a published size are left out.
What fits by category: data
CategoryFitsTightToo bigTotal
LLM222355280
Vision82321106
Image240024
Video220022
Speech400040
Embedding190019

Largest open models that fit at 4-bit

See all 409 in the table

A model fits when its 4-bit memory estimate (8k context) stays within 90% of usable memory, about 72.0 GB here. Showing the 30 largest of 409.

ModelTypeSizeMemory at 4-bit
LLM103B69.5 GB
Image83B56 GB
LLM80.4B54.2 GB
LLM78.1B51.7 GB
Vision73.4B51.4 GB
Vision72B51.3 GB
Vision71.9B51.3 GB
LLM70.6B50.3 GB
LLM70.6B50.3 GB
LLM70.6B49.5 GB
LLM70.6B49.5 GB
LLM41.8B45.3 GB
LLM67.2B45.2 GB
LLM51.6B36.8 GB
LLM51.6B36.8 GB
LLM51.6B35.2 GB
LLM51.6B35.2 GB
LLM49.1B34.7 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM36.2B26.2 GB
LLM36.2B26.2 GB
LLM35B24.9 GB
Vision33.3B24.6 GB
LLM31.2B24.1 GB
LLM32.8B24 GB
Speech33.7B24 GB

These are memory checks, not speed or runtime compatibility results.

Compared with

All hardware
HardwareTotal memoryUsable for modelsBandwidth
H100 SXMThis device
80 GBUp to 80 GB3,350 GB/s
80 GBUp to 80 GB2,039 GB/s
48 GBUp to 48 GB864 GB/s
24 GBUp to 24 GB300 GB/s

Capacity and bandwidth answer different questions: memory decides what fits, bandwidth drives how fast it runs.

Launch dates, useful-life evidence and ownership cost scenarios