Full AI value calculator

Adjust model, review share, risk premium, and monthly volume. Full TCH / TCA worksheet below (all figures per month).

Cognitive equivalency (STH)

Compare TCH (total cost of human labor) to TCA (total cost of AI): list API prices for T_in/T_out (or a fixed monthly dedicated host), plus oversight and risk. As an illustrative non-agentic anchor, ~10k tokens can resemble ~1 hour of dense reading—but task-type presets assume agentic AI (retrieval, tools, multi-step) to reflect the real world usage to increase consistency, so hourly token rates are higher.

TCA = (T_in×P_in) + (T_out×P_out) + (H_review×R_human) + C_error

All workload and cost totals are per calendar month. List prices are USD per 1M tokens; ≈ EUR uses ×0.92.

Inputs

Task type

Presets assume work is done with agentic AI (tools, retrieval, multi-turn) so API token use stays comparable across task types.

Model choice

Volume (units per month)

Task completions per calendar month (TCH, tokens, and oversight).

Human billing rate (EUR/hr)Hours per task unit

H_review (share of original task time)

C_error — risk premium (EUR / month)

TCH — human labor

Full manual cost if AI is not used. Use (?) for formulas.

Monthly workload ≈ 10 h

Workload (h/mo) = units per month × hours per unit. With your inputs: 10 × 1 h/unit ≈ 10 h/mo. This same monthly workload drives token estimates (STH × hours × volume) and oversight hours.

(10 units × 1 h/unit)

TCH (monthly)

500 €

TCH = hours per unit × EUR/hr × units per month. With your inputs: 1 × 50,00 €/hr × 10 ≈ 500,00 €/mo. Oversight (H_review × R_human) uses the same hourly rate as this TCH.

TCA — AI path

Set H_review and C_error under Inputs. Click (?) for formulas and pricing detail.

P_in / P_out (list): $2.5/1M · $15/1M USD

≈ 100,000 tokens/unit · Monthly T_in / T_out: 300,000 / 700,000

T_in and T_out come from STH × hours × monthly units, with total tokens split ~30% input / ~70% output for costing. Right now: 300,000 T_in / 700,000 T_out per month (from ≈ 100,000 tokens/unit × 10 units). List prices are USD per 1M tokens; EUR uses ×0.92 on the USD API total.

(T_in×P_in) + (T_out×P_out): $11.25≈ 10,35 €
?
(T_in × P_in) + (T_out × P_out) at list USD/1M; approximate EUR applies ×0.92 to the USD total. Using your monthly split: 300,000 T_in and 700,000 T_out → $11.25 (≈ 10,35 €).
H_review × R_human (3 h/mo): 150,00 €
?
H_review × R_human = review share × hours per unit × monthly units × your EUR/hr (from Inputs). H_review is often 20%–40% of original task time. With your inputs: 30% × 1 h/unit × 10 units × 50,00 €/hr ≈ 150,00 €/mo (3 oversight h/mo).
C_error: 0,00 €
?
C_error is the risk / hallucination premium you enter under Inputs (EUR/month). Your value: 0,00 €/mo. It adds to API or fixed cost and oversight to form TCA.
TCA (monthly): 160,35 €
?
TCA = API or fixed infra + H_review × R_human + C_error (all per calendar month). With your numbers, total 160,35 € = 10,35 € (API/fixed) + 150,00 € (oversight) + 0,00 € (C_error).

Savings

Net saving (TCH − TCA), monthly: 339,65 €
?
Net saving = TCH − TCA for the same calendar month (human-only labor cost minus full AI path cost). With your numbers: 500,00 € − 160,35 € = 339,65 €/mo.
Reduction vs TCH: 67.9%
?
Reduction vs TCH = (TCH − TCA) / TCH when TCH > 0; otherwise shown as —. With your numbers: (500,00 € − 160,35 €) ÷ 500,00 € ≈ 67.9%.

Worksheet-style check (your numbers, monthly)

Step-by-step recap of the same monthly figures as above: human cost, AI or fixed compute, oversight, and net saving — useful to sanity-check inputs. The list below updates live as you change volume, rates, hours, model, and risk inputs.

Human (TCH): 10 h × 50,00 €/hr ≈ 500,00 €
AI API: 300,000 T_in + 700,000 T_out ≈ $11.25 (≈ 10,35 € EUR)
Oversight: 3 h × 50,00 €/hr ≈ 150,00 €
Net saving:339,65 € (67.9% vs TCH)