How the stream calculator does the math

First, the calculator turns your schedule into broadcast seconds across all independently generated channels:

scheduled seconds = hours/day × 3,600 × broadcast days × channels

Then it removes the portion filled with replays:

fresh seconds = scheduled seconds × (1 − replay percentage/100)

Finally, it adds additional billable generation, such as discarded outputs or billed rerenders:

billable seconds = fresh seconds × (1 + extra percentage/100)

generation cost = billable seconds × price/second

The model uses equivalent seconds of output, not the number of API calls. A real schedule may need to round clip boundaries or leave pauses. Small differences can arise when a target duration is not divisible by the selected clip length. Check your actual output and billing records.

For one channel at 480p, two hours per day and thirty broadcast days, with no replays or extra generation:

2 × 3,600 × 30 × 1 = 216,000 new seconds

216,000 × $0.05 = $10,800 generation cost

Now fill half the show with reviewed replays and allow twenty percent extra billable generation:

216,000 × 0.5 × 1.2 × $0.05 = $6,480

The cost does not fall all the way to $5,400 because the extra generation allowance is applied to the remaining fresh output. Replay percentage describes time on air, while extra percentage describes additional billable work. They are deliberately different inputs.

A full new clip can represent one prompt slot. The calculator estimates capacity from fresh seconds only:

new slots = floor(fresh seconds ÷ seconds/clip)

paid slots = floor(new slots × paid percentage/100)

gross paid-prompt revenue = paid slots × price/paid prompt

The floor operation avoids selling fractional slots. Replays do not add new slots, and discarded generations do not add slots either. The model assumes one prompt sold per new clip; an operator with a different unit of sale needs a different revenue model.

For a continuously generated fifteen-second channel over thirty days, capacity is 172,800 slots. At twenty-five percent paid occupancy, that becomes 43,200 paid slots. This is arithmetic capacity, not a forecast of 43,200 purchases.

The payment-fee input applies a percentage to gross paid-prompt revenue. If a processor also charges a fixed amount per transaction, include the estimated monthly total in other costs. Refunds, taxes, currency conversion and bad debt are not calculated automatically.

payment fees = paid-prompt revenue × fee percentage/100

ad revenue = monthly ad impressions ÷ 1,000 × impression RPM

net result = paid-prompt revenue + ad revenue − generation cost − payment fees − other costs

Ad impressions are not viewers, pageviews or minutes watched. Use a revenue rate defined on the same impression basis. Do not enter page RPM here and assume it is interchangeable. The default advertising inputs are zero because this publication does not have evidence of your ad inventory or demand.

When at least one paid slot is selected, the calculator solves for the prompt price that would cover generation and other operating costs after percentage fees and modeled ad income:

price = max(0, generation cost + other costs − ad revenue) ÷ (paid slots × (1 − fee percentage/100))

At 768p standard pricing, a 24/7 thirty-day channel costs $207,360 in generation. With 43,200 paid slots, no advertising, no extra costs and zero payment fees, the break-even prompt price is $4.80. Payment fees and operating costs raise that threshold.

If no paid slots are selected, the calculator displays an unavailable break-even price rather than dividing by zero. If modeled ad revenue already covers the modeled costs, the minimum additional prompt price is zero. That result still depends entirely on the entered ad assumptions.

The calculator does not infer audience size, retention, conversion, inventory fill or sponsorship terms. It does not automatically price storage, video delivery, staff, support, moderation, chargebacks or taxes. The “other costs” field exists so you can add a known monthly allowance.

Price presets are dated public references. Standard H3 Max rates are used by default. Launch discounts are available as a scenario because official pages describe the promotional window differently; they are not an assurance that your account can claim a discount today.

The interface validates negative, non-finite and out-of-range inputs. A shared URL contains numeric scenario settings, not prompts, email addresses or provider keys. CSV exports retain both inputs and unrounded calculation results, making it easier to check a budget outside the website.

Start with a conservative schedule, zero unproven revenue and measured costs. Then compare a shorter show, a replay-heavy show and a full-time channel. Change one variable at a time so you can see what drives the difference.

Once a real pilot runs, replace assumptions with observed billable seconds, accepted output, payment charges and delivery costs. A negative modeled result is useful information; it may point toward a smaller schedule, a different product or a decision not to operate continuously.

Neither a calculator result nor an API benchmark establishes profitability. The builder guide connects the budget to operational controls, including spend reservations and a stop switch.