Skip to content

Self-Host vs API Break-Even Calculator

GPU rental, ops time and real utilisation against per-token API spend, solved for the monthly request volume at which running your own box starts to win.

The workload, and what the API charges for it

The box you would run instead

Self-hosting starts winning above
4,037,037 requests / month

You are at 2,000,000 — 49.5% of the way there. The API costs $2,700 a month at that volume; the box costs $5,450 whether you send it one request or a million.

API cost per request
$0.0014
API cost per month
$2,700
GPU rental per month
$3,650
Ops time per month
$1,800
Self-hosted cost per month
$5,450
Capacity at 100% utilisation
4,730M output tokens / mo
Capacity at 35.0% utilisation
1,656M output tokens / mo
Output tokens you actually need
800M / mo
Utilisation this workload implies
16.9%
GPUs needed at your utilisation
2 is enough
Self-hosted, per 1M output tokens at 35.0%
$3.29
Self-hosted, per 1M output tokens at 100%
$1.15
API, per 1M output tokens (blended)
$3.38
Difference per month at your volume
$2,750
What this assumes: the rented GPU is billed for every hour it is up, used or not — that is the whole reason utilisation dominates this calculation. Throughput is measured in output tokens per second, because decode is the binding constraint on a serving box; prefill is far faster per token and is ignored here, which flatters the self-hosted side. The API figure is blended across input and output at your two rates, then divided by output tokens so the two $/1M numbers are comparable on the same denominator. Ops time is a real cost and is included; the engineering weeks to get the first version serving are not. Nor are egress, storage for weights, a staging environment, or the second box you need before you can take the first one down for an upgrade.

Utilisation is the term that decides this and it is the term most published comparisons omit. The arithmetic people usually run divides a GPU's hourly rate by its peak throughput, arrives at a cost per million tokens far below any API rate, and concludes that self-hosting is obviously cheaper. It would be, if the GPU generated tokens every second of every hour you paid for it. It does not. Traffic is diurnal, traffic is bursty, and you have to size for the peak, which means the trough is idle capacity you are still renting.

Run the sensitivity yourself: halve the utilisation field and watch the self-hosted cost per million double. Nothing else on this page has that leverage. A twenty per cent improvement in throughput moves the answer by twenty per cent; going from 30% utilisation to 60% halves your cost per token. That is why the serious version of this decision is a scheduling problem — filling the troughs with batch work, consolidating several workloads onto one fleet — rather than a hardware-selection problem.

The break-even is expressed as a monthly request count because that is the number you can check against your own dashboard tomorrow. Below it, an API is cheaper and also carries none of the risk. Above it, the box is cheaper on paper, and the remaining question is whether the ops burden — which is included here as a line item and is routinely underestimated — is one your team wants. There is a wide band around the crossing point where the honest answer is that the money is a wash and you should decide on other grounds: data residency, model choice, latency floor, or simply which failure you would rather be woken up for.

Self-Host vs API Break-Even Calculator · Multigrid