NVIDIA / discrete

NVIDIA H100 80GB SXM

80 GB decides what fits. 3350 GB/s decides how fast it runs once it does.

Quick answer

The NVIDIA H100 80GB SXM fits 244 of 320 sized models entirely in its 80 GB; the largest widely used one is Ling-3.0-flash (78.2 GB at Q4_K_M). Memory decides what fits; its 3350 GB/s of bandwidth decides how fast it answers.

Largest popular model that fitsLing-3.0-flash127B params · Q4_K_M · needs 78.2 GB~745 tok/s Faster than you readOpen in the calculator →
Most downloaded that fitsQwen3.8-27B28B params · Q4_K_M · needs 18.6 GB~154 tok/s Faster than you readOpen in the calculator →

At 8,192 tokens of context with the whole model in device memory. Speeds are estimates from memory bandwidth, not benchmarks run on this card.

NVIDIA H100 80GB SXM · 80 GB · 3350 GB/s · Q4_K_M where published · 8,192 tokensSpecificationpublished dataDecode speedestimate

244 of 320 fit entirely

244 of the 320 models the engine can size fit entirely in device memory at Q4_K_M where it is published, otherwise the nearest published format, and 8,192 tokens. 2 run with some layers on system memory (32 GB assumed), and 74 do not run at all.

76%of the models the engine can size fit entirely

80GB of device memory

3350GB/s memory bandwidth

2run with system memory

74do not run at 8,192 tokens

Featured models · Q4_K_M at 8,192 tokens, including offload

Model	Needs	Verdict	Decode	Calculator
Qwen3.5-2B2.3B parameters	2.32 GB	fits	~1847 tok/s	Open →
NVIDIA-Nemotron-3-Nano-4B-BF164.0B parameters	3.46 GB	fits	~856 tok/s	Open →
Qwen3-VL-8B-Instruct8.8B parameters	7.04 GB	fits	~387 tok/s	Open →
Qwen3.5-9B9.7B parameters	7.05 GB	fits	~469 tok/s	Open →
gemma-4-12B-it12B parameters	9.54 GB	fits	~279 tok/s	Open →
gpt-oss-20b21B parameters	13.8 GB	fits	~911 tok/s	Open →
Qwen3.8-27B28B parameters	18.6 GB	fits	~154 tok/s	Open →
Qwen3.6-27B28B parameters	18.6 GB	fits	~154 tok/s	Open →
Kimi-Linear-48B-A3B-Instruct49B parameters	30.4 GB	fits	~1218 tok/s	Open →
Qwen3.8-Flash-Next180B parameters	111.6 GB	needs 32.2 GB of system RAM; 32 GB assumed	does not run	Open →
DeepSeek-V4-Flash-Vision-Exp305B parameters	188.1 GB	needs 108.9 GB of system RAM; 32 GB assumed	does not run	Open →
GLM-5.3-Flash321B parameters	198.3 GB	needs 118.4 GB of system RAM; 32 GB assumed	does not run	Open →

These examples are selected from prominent labs using the catalogue’s latest Hugging Face 30-day downloads and repository-creation freshness signal, with newer releases guaranteed a place. Offloaded rows assume 32 GB of system RAM, and a speed is only shown for a row that runs. Fitting in memory is not the same as loading: whether the runtime and version you have supports each architecture and format on this machine has not been tested here. Each “Open” link carries the same model, format, context and RAM into the calculator.

Fits entirely in NVIDIA H100 80GB SXM memory — 244 of 320 sized models

Every model that fitswhole model resident · offload off

The most demanding model that fits is Devstral-2-123B-Instruct-2512 at Q4_K_M: 78.6 GB of the 80 GB, leaving 1.35 GB spare.

Fully resident at 8,192 tokens (or the model’s own maximum, where that is shorter), offload off, at 3350 GB/s. Each row is a run of the engine for this configuration; the rows start with current, prominent releases and “fits” is memory, not a tested runtime. Older or less prominent models remain available through this search and “Show all”.

Find a model in this list

Showing 40 of 244 models that fit.

Model	Format	Needs	Spare	Decode	Downloads / 30d	Calculator
Qwen3.8-27BQwen · 28B params	Q4_K_M	18.6 GB	61.4 GB	~154 tok/s	6.9M	Open →
gemma-4-26B-A4B-itGoogle · 26B params	Q4_K_M	16.8 GB	63.2 GB	~856 tok/s	13M	Open →
gemma-4-31B-itGoogle · 31B params	Q4_K_M	22.2 GB	57.8 GB	~111 tok/s	9.9M	Open →
Qwen3.5-9BQwen · 9.7B params	Q4_K_M	7.05 GB	73.0 GB	~469 tok/s	9M	Open →
Qwen3.5-4BQwen · 4.7B params	Q4_K_M	3.98 GB	76.0 GB	~843 tok/s	7.8M	Open →
Qwen3.6-35B-A3BQwen · 36B params	Q4_K_M	23.1 GB	56.9 GB	~1229 tok/s	3.3M	Open →
gemma-4-12B-itGoogle · 12B params	Q4_K_M	9.54 GB	70.5 GB	~279 tok/s	1.9M	Open →
Qwen3.6-27BQwen · 28B params	Q4_K_M	18.6 GB	61.4 GB	~154 tok/s	2.5M	Open →
Qwen3.5-2BQwen · 2.3B params	Q4_K_M	2.32 GB	77.7 GB	~1847 tok/s	4.9M	Open →
gemma-4-E4B-itGoogle · 8.0B params	Q4_K_M	5.86 GB	74.1 GB	~753 tok/s	4.4M	Open →
NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA · 32B params	Q4_K_M	20.3 GB	59.7 GB	~117 tok/s	530K	Open →
NVIDIA-Nemotron-3-Nano-4B-BF16NVIDIA · 4.0B params	Q4_K_M	3.46 GB	76.5 GB	~856 tok/s	3.4M	Open →
gemma-4-E2B-itGoogle · 5.1B params	Q4_K_M	4.02 GB	76.0 GB	~1989 tok/s	3M	Open →
Qwen3.5-0.8BQwen · 873M params	Q4_K_M	1.46 GB	78.5 GB	~4045 tok/s	2.6M	Open →
Qwen3-VL-8B-InstructQwen · 8.8B params	Q4_K_M	7.04 GB	73.0 GB	~387 tok/s	14.6M	Open →
Qwen3.5-27BQwen · 28B params	Q4_K_M	18.6 GB	61.4 GB	~154 tok/s	1.9M	Open →
North-Micro-Vision-InstructCohereLabs · 2.5B params	Q4_K_M	2.99 GB	77.0 GB	~1214 tok/s	180.7K	Open →
Qwen3.5-35B-A3BQwen · 36B params	Q4_K_M	23.1 GB	56.9 GB	~1229 tok/s	1.6M	Open →
NVIDIA-Nemotron-3-Super-120B-A12B-BF16NVIDIA · 124B params	Q4_K_M	76.9 GB	3.08 GB	~30 tok/sAbout reading pace	1.2M	Open →
granite-4.2-8bIBM · 8.8B params	Q4_K_M	7.49 GB	72.5 GB	~353 tok/s	128.4K	Open →
GLM-4.7-Flashzai-org · 31B params	Q4_K_M	20.4 GB	59.6 GB	~116 tok/s	1.8M	Open →
granite-4.1-3bIBM · 3.4B params	Q4_K_M	3.72 GB	76.3 GB	~822 tok/s	521.8K	Open →
LFM2.5-2.6BLiquidAI · 2.7B params	Q4_K_M	2.61 GB	77.4 GB	~1319 tok/s	108.1K	Open →
Qwen3-VL-4B-InstructQwen · 4.4B params	Q4_K_M	4.72 GB	75.3 GB	~615 tok/s	3.5M	Open →
granite-4.1-30bIBM · 29B params	Q4_K_M	20.4 GB	59.6 GB	~116 tok/s	301.5K	Open →
Qwen3.5-122B-A10BQwen · 125B params	Q4_K_M	77.9 GB	2.05 GB	~450 tok/s	512.6K	Open →
Qwen3-VL-2B-InstructQwen · 2.1B params	Q4_K_M	3.05 GB	77.0 GB	~1115 tok/s	2.8M	Open →
granite-4.2-3bIBM · 3.7B params	Q4_K_M	3.72 GB	76.3 GB	~822 tok/s	50.6K	Open →
LFM2.5-230MLiquidAI · 230M params	Q4_K_M	1.05 GB	78.9 GB	~9194 tok/s	88.6K	Open →
Qwen3-0.6BQwen · 752M params	Q4_K_M	2.22 GB	77.8 GB	~1711 tok/s	29.7M	Open →
Qwen3-Coder-NextQwen · 80B params	Q4_K_M	50.0 GB	30.0 GB	~817 tok/s	596.3K	Open →
gpt-oss-20bOpenAI · 21B params	Q4_K_M	13.8 GB	66.2 GB	~911 tok/s	6.6M	Open →
granite-4.2-30bIBM · 29B params	Q4_K_M	20.7 GB	59.3 GB	~116 tok/s	35.8K	Open →
granite-4.1-8bIBM · 8.8B params	Q4_K_M	7.49 GB	72.5 GB	~353 tok/s	179.3K	Open →
NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA · 32B params	Q4_K_M	20.3 GB	59.7 GB	~117 tok/s	875.2K	Open →
gpt-oss-120bOpenAI · 117B params	Q4_K_M	71.9 GB	8.06 GB	~641 tok/s	4.5M	Open →
LFM2.5-VL-3BLiquidAI · 3.1B params	Q4_K_M	2.85 GB	77.1 GB	~1319 tok/s	26.7K	Open →
MiMo-V2.6-Distill-Qwen-9BNEWXiaomiMiMo · 9.4B params	Q4_K_M	6.90 GB	73.1 GB	~469 tok/s	14.6K	Open →
Qwen3-4B-Instruct-2507Qwen · 4.0B params	Q4_K_M	4.72 GB	75.3 GB	~615 tok/s	3.7M	Open →
Ling-3.0-tinyinclusionAI · 7.9B params	Q4_K_M	5.85 GB	74.1 GB	~2656 tok/s	17.7K	Open →

Sized at the format most people actually download, not at FP16. The catalogue ordered by downloads →

Rent it instead

Rent an H100 SXM by the hour

What the NVIDIA H100 80GB SXM costs on Vast.ai and RunPod right now, per machine, from one card to eight. Every price is read from the provider's own API.

Prices · 18:30 UTC, 22 Sept

1× H100 SXM · 80 GB

$1.74per hour

Vast.ai · Marketplace · 98.7% reliableRunPod $3.49/h Secure CloudRent on Vast.ai

2× H100 SXM · 160 GB

$3.47per hour

Vast.ai · Marketplace · 98.7% reliableRunPod $6.98/h Secure CloudRent on Vast.ai

4× H100 SXM · 320 GB

$9.07per hour

Vast.ai · Marketplace · 99.9% reliableRunPod $13.96/h Secure CloudRent on Vast.ai

Buy one or rent one?

Your price, your hours, your electricity. Everything else is arithmetic.

Price you would pay ($)

Resale after 3 years ($)

Hours of use per day4 h

$ per kWh

Watts

Rent $/h

Watts start at the NVIDIA H100 80GB SXM’s published board power (700 W), an upper bound: decoding rarely holds a card at its limit. Rent starts at the cheapest live price (Vast.ai). The electricity price is a placeholder — put yours in.

At 4 h a day, renting costs $2,540.40 a year.

That is $211.70 a month, and nothing when the machine is stopped.
Owning would cost $255.50 a year in electricity, on top of the price.
Every $1,000 of purchase price takes 0.4 years of this use to earn back.

Enter the price you would pay to see the exact break-even.

On-demand prices for the whole machine. Vast hosts below 98% measured reliability are left out; RunPod’s Community Cloud is vetted third-party hosts and its Secure Cloud is data-centre capacity. Every rentable card compared.

Referral links Vast.ai, RunPod and Novita pay us a share of what you spend if you sign up through these buttons. It costs you nothing, and it never decides an order or a recommendation: both are computed from the live price and the speed, and options that pay us nothing are listed and recommended on the same terms. How we rank

Under the hood

The specification behind every figure

What the manufacturer publishes for this device, and the pages it was read from.

Manufacturer specification

Memory scope	dedicated
Capacity	80 GB
Published options	80 GB
Bandwidth	3350 GB/s
Memory type	HBM3
Bus width	5120-bit
FP32 peak	67 TFLOPS
Dense matrix peak	Not published without sparsity
Max thermal design power (TDP), configurable ceiling	700 W

Source ledger

NVIDIA H100 GPU ↗

Retrieved 2026-08-30

NVIDIA H100 Tensor Core GPU Architecture ↗

Retrieved 2026-08-30

Caveats

The current product page supersedes the whitepaper's older/pre-final 3352GB/sec and 66.9 non-Tensor TFLOPS entries. See conflicts.
No dense matrix throughput is catalogued for this device: NVIDIA's current H100 pages publish FP16 Tensor Core throughput only with sparsity. Time to first token is withheld rather than priced against a sparsity figure.

Published capacity is a hardware ceiling, not guaranteed free runtime memory. The calculator shows the runtime reserve separately rather than folding it into a single number.

Run the diagnostic on the NVIDIA H100 80GB SXM →