CalcSays
NVIDIA A100 80GB · SELF-HOST BREAK-EVEN

NVIDIA A100 80GB vs API Break-Even

Renting NVIDIA A100 80GB costs $1.49/hour (RunPod) — a fixed monthly fee. This page computes exactly how many requests/month it takes to beat paying per API token.

For teams deciding whether to rent GPUs and self-host an open-weight model — computes the exact monthly request-volume break-even against the same model's API price, not a 'self-hosting is cheaper at scale' assumption.

Model prices from OpenRouter · updated 2026-07-13

01 Fixed rental vs variable API

GPU

Compared against Qwen3 32B's live API price. Not modeled: whether this GPU can actually serve this many requests fast enough — see the FAQ below.

02 Fixed vs variable, at this volume

API still wins — self-hosting costs 2.1× the API bill.

GPU rental (fixed)
$1,088
1× $1.49/hr × 730hr
API (variable)
$528
$0.000176/request

Break-even: 6,180,114 requests/month

Tweet
Why does the GPU cost the same in a quiet month? Rental bills by the hour of availability, not by usage — you pay for all 730 hours whether the GPU serves one request or ten million. The API has no such floor.
📄 Get the full agent cost audit
A multi-page report on your exact setup: where cost spirals, how much context-trimming / state-resets / model-routing each saves, and which model+framework combo is cheapest at your real failure rate. Generated from your numbers.

Inference prices from OpenRouter, snapshot 2026-07-13, synced daily. GPU rental prices hand-verified (checked 2026-07-13), re-audited quarterly — they move faster than API prices: RunPod GPU Cloud pricing. All math runs in your browser.

How the math works

NVIDIA A100 80GB rents at $1.49/hour on-demand (RunPod, checked 2026-07-13). Rented by the hour means it bills for all 730 hours in a month — 1 × $1.49 × 730 = $1,088/month, whether the GPU serves one request or ten million.

Qwen3 32B's API price is the opposite shape: purely variable. At 800 input + 400 output tokens per request, each call costs $0.000176 ($0.000064 in + $0.000112 out). At 3,000,000 requests/month that's $528 — scaling down to near-zero in a quiet month, unlike the GPU rental.

Same baseline, one identity: API cost ($528/mo) minus GPU rental ($1,088/mo) = −$560. At this volume, the API still wins by $560/month — the GPU's fixed rental costs more than the traffic justifies.

The number that actually matters isn't "which is cheaper per token" — it's the break-even: 6,180,114 requests/month. Below it, the API's pay-per-use scaling is cheaper; above it, the GPU's flat monthly fee wins, and every request past that point costs nothing extra (until you outgrow the GPU's throughput).

Not modeled: whether NVIDIA A100 80GB can actually serve this many requests fast enough — throughput depends on model size, batch size, and sequence length in ways no flat number captures honestly. This page compares cost only; check the GPU's real throughput for your model before committing. GPU rates hand-verified (checked 2026-07-13), re-audited quarterly — rental prices move faster than API prices. Inference prices sync daily from OpenRouter (updated 2026-07-13).

Frequently asked questions

Should a NVIDIA A100 80GB deployment self-host on NVIDIA A100 80GB?

At this page's defaults — 3,000,000 requests/month, 800/400 in/out tokens — not yet: the API is still $560/month cheaper — the GPU rental costs more than the traffic justifies. Tune the sliders for your real numbers.

How many requests per month until self-hosting pays off?

6,180,114 requests/month — the point where the API's variable cost equals the GPU's fixed monthly rental ($1,088). Past that volume, every extra request costs nothing extra on the GPU side.

Why does the GPU cost the same even in a quiet month?

Because GPU rental bills by the hour of availability, not by usage — NVIDIA A100 80GB at $1.49/hr costs $1,088 for the month regardless of whether it served 10 requests or 10 million. The API has no such floor: it scales down with traffic.

What does NVIDIA A100 80GB cost to rent?

$1.49/hour on-demand from RunPod (checked 2026-07-13). At 1 GPU × 730 hours/month, that's $1,088 — a fixed monthly cost, not per-request.

Does this account for whether the GPU can handle the traffic?

No — this page compares cost only. Whether a given GPU can actually serve this many requests per month depends on model size, batching strategy, and sequence length, none of which have a single honest flat number. Check the GPU's real throughput for your specific model before committing to a break-even volume.

Are these prices current?

GPU rental rates are hand-verified against RunPod's pricing page (checked 2026-07-13) and re-audited quarterly — GPU prices move faster than LLM API prices, so treat this as a snapshot, not a guarantee. Inference prices sync daily from OpenRouter (updated 2026-07-13). All math runs client-side with tested code.