Compute earnings estimator

What could your compute earn?

A quick, directional estimate of what your CPU/GPU capacity could earn as governed, evidence-backed compute under HUMENS Compute-as-a-Service Virtual Aggregation Plant (CaaS-VAP, bring your hardware) or Compute-as-a-Service Virtual Power Plant (CaaS-VPP, energy-backed). The real number comes from a governed evaluation — this is a starting point.

Your capacity

Add your machines

Availability (how often it's free to work)

Sellable hours / month

Usable share of capacity: 50%

Loading model…

Estimated governed compute revenue to you

$0 – $0/mo

$0 – $0 / year

Your share of governed compute revenue, before any energy or infrastructure costs, shown as a range because utilization and workload mix vary. CaaS-VAP is pure revenue-share on compute you own; CaaS-VPP (energy-backed, behind-the-meter) is quoted separately during your evaluation. HUMENS margin / RevSplit is never shown on this public page.

Illustrative token throughput (AI factory output)

0 tokens/sec

0 tokens/sec at 50% · 0 tokens/mo

Cores → energy → tokens. GPU-dominated, illustrative projection against a stated reference model — never measured or guaranteed throughput.

Illustrative behind-the-meter energy (at your utilization)

0 kW

Derived from your hardware's nameplate power draw × utilization — the behind-the-meter power this compute activates at the edge. Access, not central generation, and no grid-export or wholesale claim. Metered, evidence-backed figures come from a governed evaluation.

Book an Evaluation Or call (844) HUMENS-1

Want the governed, evidence-backed number for your exact hardware and energy profile? A HUMENS operator runs the full Compute-Capacity Unit (CCU) evaluation with you — no obligation.

How to read your estimate

The three numbers, defined

Estimated governed compute revenue: Your supplier share of governed compute revenue, shown as a range because utilization and workload mix vary. It is before energy and infrastructure costs, and before HUMENS' share. Illustrative — not a quote or a guarantee of income.
Illustrative token throughput (AI-factory output): What your hardware could produce — tokens/sec at full load and at your utilization, plus tokens per kWh (intelligence per unit of energy). Tokens are GPU-dominated: CPU cores add very little, so a large CPU-only core count is not large token output. These are projections against a stated reference model/precision, never measured or guaranteed throughput.
Behind-the-meter (BTM) energy potential: The power your hardware draws and activates at the edge — derived directly from its nameplate power draw × utilization (potential at 100%, engaged at your slider, and MWh/month over your sellable hours). This is the energy your machines consume where they sit, behind your meter — not power we generate, and not a grid-export or wholesale claim.

Not the same as a fleet energy number. In a governed CaaS-VPP evaluation we separately measure your site's accessible headroom (spare service capacity you could divert) and aggregate it across locations. That is a distinct, metered figure — this estimator only shows what the hardware you entered would draw.