The same task can end up costing three times more just because the wrong billing model was picked. Residential and mobile traffic is billed per gigabyte; datacenter and ISP ports are billed by rental term. Without running the numbers it is easy to overpay either way. Here is the break-even point and how to size the spend for a specific task.

How each model is calculated

Traffic billing is a price per gigabyte transferred; the address is issued per session and rotates on timeout or on request. Spend grows linearly with task volume: 500 MB today, nothing tomorrow — the bill equals the sum actually transferred.

Dedicated port billing is a fixed price for a specific IP for a term: a day, a week, a month. Traffic within that term is not billed separately — push 50 GB or 500 MB through the same address, price stays the same. Spend depends on the number of ports and the term, not on data volume.

Comparing the two on real volumes

Typical figures: a text catalogue page without media is 0.3–1 MB, a product card with images is 2–5 MB. A thousand product-card checks fit into 3–5 GB, a hundred into 0.3–0.5 GB.

For a light task — a hundred cards a day, up to 0.5 GB — traffic billing is almost always cheaper: you pay for megabytes actually transferred instead of holding a port idle most of the day. For a heavier task — several thousand checks a day over a month — cumulative traffic easily runs into hundreds of gigabytes, and the price per gigabyte times that volume starts to exceed a dedicated port's fixed monthly cost.

The break-even point

The break-even point is the monthly traffic volume at which a gigabyte package costs the same as a month-long port rental: the port's monthly price divided by the price per gigabyte gives a gigabyte-equivalent figure. Below that number, paying for traffic is cheaper. Above it, a dedicated port pushing any volume through without extra charges wins.

For datacenter and ISP ports, break-even is usually reached at moderate volumes because the port is cheap. For residential and mobile traffic the threshold is higher, but holding the same address indefinitely is physically limited — the pool keeps rotating, so a long-term fixed IP is not really the same offer as a datacenter port. The choice hinges on address nature too, covered in mobile versus residential proxies.

Which scenario favors which model

A gigabyte package is cheaper for one-off checks — a competitor price scan, an occasional export, testing a script before a production run — where a port would sit idle most of the time while traffic billing pays for what gets used. It also wins on unpredictable load: if volume swings from 200 MB to 20 GB by the day, a fixed port means overpaying on quiet days, while a package follows actual load automatically. And it fits cases where address reputation matters more than keeping the same IP — a marketplace storefront check, an antidetect browser — since changing address between sessions is not a problem and often keeps each check clean.

A dedicated port wins under constant background load: a server querying an API or serving traffic all day easily runs past break-even within a month. Same for a long-lived account tied to one IP, where rotating the address is a problem, not an option. Bulk data collection follows the same logic: the more gigabytes pass through per month, the faster the port's fixed price pays for itself against a linearly growing bill.

How to estimate your own spend in advance

The formula: operations per day × average weight per operation × days in the month = spend in gigabytes. Average weight comes from the content type — text page, image-heavy card, video feed — refined by a real test run rather than table figures: picking the cheapest option blindly is a reliable way to miss the budget. Compare the result to the channel's break-even point for a clear answer on which model to pick. Also check address quality — an unstable channel drives up spend through retries either way.

Frequently Asked Questions

Can the break-even point be calculated before a first test run?

Roughly, yes, using table reference figures for page weight. The exact number only emerges after a real run, because page weight depends on the specific site — media and scripts differ several times over between platforms.

What happens if a traffic package runs out before the term ends?

Top-ups work the same way, by the gigabyte, without recalculating what has already been spent. The task does not stop as long as auto-renewal is enabled or there is a balance to draw from.

Can both models be combined within one task?

Yes, that is standard practice: constant background load stays on a dedicated port, while occasional or geo-sensitive checks are covered by a traffic package, so each part of the task pays under its own model.

Both billing models, with the cost worked out for a specific volume, are available in the proxy section: the product card shows right away whether it is a rental term or a gigabyte package.