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

GPT-OSS 120B (MoE) on H100 SXM5 80GB: Does It Fit, How Fast, What It Costs

GPT-OSS 120B (MoE) at AWQ 4-bit needs 64.3 GB for its weights and 66.7 GB in total at 8K context with 8 concurrent users. That means it fits on a single H100 SXM5 80GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
66.7 GB needed / 80 GB available
Decode speed
683.1
tok/s per stream
Aggregate
8198
tok/s @ 8 concurrent
Cost / 1M tokens
$0.0847
1× 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 GB66.7 GB683.18198200.3 ms$0.0847✓
2× H100160 GB66.7 GB116127872115.3 ms$0.0498✓
4× H100320 GB66.7 GB232311148872.6 ms$0.0249✓
8× H100640 GB66.7 GB464544595251.3 ms$0.0125✓
Concurrency vs context length
On 1× H100 SXM5 80GB at AWQ 4-bit. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K101309.2 MB✓ fits
8K51618.5 MB✓ fits
33K122.42 GB✓ fits
66K64.83 GB✓ fits
131K39.66 GB✓ fits
Other GPUs that run GPT-OSS 120B (MoE)
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB4×59638$0.009316
AMD MI300X 192GB1×12969$0.0535
L40S 48GB2×7188$0.0541
AMD MI325X 288GB1×14682$0.0662
A100 80GB SXM41×4894$0.0681
Intel Gaudi 3 128GB1×8075$0.0688
Frequently asked questions

Can GPT-OSS 120B (MoE) run on H100 SXM5 80GB?

Yes. GPT-OSS 120B (MoE) at AWQ 4-bit needs 66.7 GB including the KV cache, and H100 SXM5 80GB has 80 GB — so it fits on a single GPU.

How fast is GPT-OSS 120B (MoE) on H100 SXM5 80GB?

683.1 tokens/sec per stream and 8198 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 200.3 ms, and inter-token latency is 1.5 ms.

What does GPT-OSS 120B (MoE) on H100 SXM5 80GB cost per million tokens?

$0.0847 per 1M output tokens at 1× 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 GPT-OSS 120B (MoE) need on H100 SXM5 80GB?

64.3 GB for the weights at AWQ 4-bit, plus 2.5 GB of KV cache for 8 concurrent requests at 8K context — 66.7 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is H100 SXM5 80GB the right choice for GPT-OSS 120B (MoE)?

It works. Compare $0.0847 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 8× RTX 3080 10GB at $0.004393, and the fastest is 2× Cerebras CS-3 (WSE-3) at 166400 tok/s.

Adjust the workload yourself

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

Open in calculator