NVIDIA80 GB VRAM3,350 GB/s$2.5/hr

Qwen 2.5-VL 72B (VLM) on H100 SXM5 80GB: Does It Fit, How Fast, What It Costs

Qwen 2.5-VL 72B (VLM) at FP8 (E4M3) needs 72.0 GB for its weights and 93.5 GB in total at 8K context with 8 concurrent users. That means it needs 2× H100 SXM5 80GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 2×
93.5 GB needed / 160 GB available
Decode speed
77.12
tok/s per stream
Aggregate
1851
tok/s @ 8 concurrent
Cost / 1M tokens
$0.7504
2× H100 SXM5 80GB
What changes as you add H100 SXM5 80GBs
Tensor-parallel degree vs VRAM, throughput, and cost. Cost scales linearly with GPU count; decode throughput scales near-linearly too, so cost per token stays roughly flat.
GPUsTotal VRAMVRAM neededDecode tok/sAggregate tok/sTTFTCost / 1M tokFits?
1× H10080 GB93.5 GB0.0000.000——✗
2× H100160 GB93.5 GB77.1218511.23 s$0.7504✓
4× H100320 GB93.5 GB154.27403632.3 ms$0.3752✓
8× H100640 GB93.5 GB308.529614331.1 ms$0.1876✓
Concurrency vs context length
On 2× H100 SXM5 80GB at FP8 (E4M3). KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K651.34 GB✓ fits
8K322.68 GB✓ fits
33K810.74 GB✓ fits
66K421.47 GB✓ fits
131K242.95 GB✓ fits
Other GPUs that run Qwen 2.5-VL 72B (VLM)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB4×3960$0.1403
A100 80GB SXM42×1105$0.6033
AMD MI300X 192GB1×861.3$0.8063
L40S 48GB2×477.4$0.8147
AMD MI325X 288GB1×975.0$0.9972
Intel Gaudi 3 128GB1×536.3$1.04
Frequently asked questions

Can Qwen 2.5-VL 72B (VLM) run on H100 SXM5 80GB?

Yes, but it needs 2× H100 SXM5 80GB. At FP8 (E4M3) the total footprint is 93.5 GB (72.0 GB of weights plus 21.5 GB of KV cache at 8K × 8 concurrent users), against 160 GB of available VRAM across 2 GPUs.

How fast is Qwen 2.5-VL 72B (VLM) on H100 SXM5 80GB?

77.12 tokens/sec per stream and 1851 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 1.23 s, and inter-token latency is 13.0 ms.

What does Qwen 2.5-VL 72B (VLM) on H100 SXM5 80GB cost per million tokens?

$0.7504 per 1M output tokens at 2× H100 SXM5 80GB ($2.50/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Qwen 2.5-VL 72B (VLM) need on H100 SXM5 80GB?

72.0 GB for the weights at FP8 (E4M3), plus 21.5 GB of KV cache for 8 concurrent requests at 8K context — 93.5 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is H100 SXM5 80GB the right choice for Qwen 2.5-VL 72B (VLM)?

It works. Compare $0.7504 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 4080 16GB at $0.1052, and the fastest is 4× Cerebras CS-3 (WSE-3) at 44200 tok/s.

Adjust the workload yourself

Same formulas, your context length, concurrency, and quantisation.

Open in calculator