Most data centers are underutilized. This calculator estimates the incremental revenue from selling that idle capacity through HZ AI. Adjust the inputs to match your site.
Rates above are editable, the defaults are July 2026 wholesale market anchors: H100 $1.85 · H200 $2.50 · B200 $4.25 · B300 $5.75 per GPU-hour.
Assumed sustained output tok/s per GPU (70B-class, batched, pre-benchmark ballparks): H100 600 · H200 850 · B200 1,300 · B300 1,500.
| Line | Site / mo | Site / yr |
|---|---|---|
| Incremental GPU-hours sold | ||
| Tokens billed (input + output) | ||
| Data center earnings |
Earnings shown are the data center's revenue share minus HZ AI's platform fee. Assumes 730 hours per month. Earnings require matched buyer demand and committed (reserved) volume makes these numbers contractual rather than best-case.
Token mode assumes the per-class tokens/sec figures are sustained across all sold hours. Those are published-benchmark ballparks until HZ AI's own throughput benchmark lands. Quote GPU-hour pricing where a guarantee is needed.
Prices are July 2026 market anchors and move with supply. Blackwell-class rates are still settling. This is an estimate only and not a guarantee.