NVIDIA80 GB VRAM2,000 GB/s$1.2/hr

GPT-OSS 120B (MoE) on A100 80GB SXM4: 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 A100 80GB SXM4. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
66.7 GB needed / 80 GB available
Decode speed
407.8
tok/s per stream
Aggregate
4894
tok/s @ 8 concurrent
Cost / 1M tokens
$0.0681
1× A100 80GB SXM4
What changes as you add A100 80GB SXM4s
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× A10080 GB66.7 GB407.84894568.1 ms$0.0681✓
2× A100160 GB66.7 GB693.316640299.3 ms$0.0401✓
4× A100320 GB66.7 GB138766560164.6 ms$0.0200✓
8× A100640 GB66.7 GB277326624097.3 ms$0.0100✓
Concurrency vs context length
On 1× A100 80GB SXM4 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
Intel Gaudi 3 128GB1×8075$0.0688
H100 SXM5 80GB1×8198$0.0847
Frequently asked questions

Can GPT-OSS 120B (MoE) run on A100 80GB SXM4?

Yes. GPT-OSS 120B (MoE) at AWQ 4-bit needs 66.7 GB including the KV cache, and A100 80GB SXM4 has 80 GB — so it fits on a single GPU.

How fast is GPT-OSS 120B (MoE) on A100 80GB SXM4?

407.8 tokens/sec per stream and 4894 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 568.1 ms, and inter-token latency is 2.5 ms.

What does GPT-OSS 120B (MoE) on A100 80GB SXM4 cost per million tokens?

$0.0681 per 1M output tokens at 1× A100 80GB SXM4 ($1.20/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 A100 80GB SXM4?

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 A100 80GB SXM4 the right choice for GPT-OSS 120B (MoE)?

It works. Compare $0.0681 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