Pay-per-GB vs Pay-per-IP: Which Proxy Payment Model is Best for Your Business?

When choosing a proxy, you will inevitably encounter two payment models: pay-per-GB (payment for traffic) and pay-per-IP (payment for the number of IP addresses). A wrong choice can increase your costs by 3-5 times, especially if you are working with multi-accounting, arbitrage, or scraping. In this article, we will analyze both models with practical calculations for different tasks and show when each model is more profitable.

What are pay-per-GB and pay-per-IP: main differences

The proxy payment model determines what exactly you are paying for: the volume of data transmitted or the number of IP addresses at your disposal. These are fundamentally different approaches that suit different types of tasks.

Pay-per-GB (payment for traffic) — you purchase a certain volume of data, for example, 10 GB or 100 GB. The provider gives you access to a pool of IP addresses (usually thousands or tens of thousands of IPs), and you use as many IPs as needed until you exhaust the purchased traffic. This model is popular among residential proxy providers, where access to a large number of IPs from different regions is important.

Pay-per-IP (payment for IP addresses) — you pay for a specific number of IP addresses for a certain period, usually a month. Each IP is assigned specifically to you, and you can use unlimited traffic through these IPs. This model is typical for mobile proxies and data center proxies, where stability and consistency of the IP address are important.

Pay-per-GB: when payment for traffic is beneficial

The pay-per-GB model is ideal for tasks where you need access to a large number of IP addresses from different regions, but the volume of traffic is relatively small. This is typical for working with anti-detect browsers, where each account uses its own IP, but the actual traffic is minimal — page loading, form submissions, viewing ad accounts.

Typical use cases for pay-per-GB:

• Multi-accounting on social networks and advertising platforms — working with 20-100 accounts in Dolphin Anty, AdsPower, or Multilogin. Each account uses 50-200 MB per month, which gives 5-10 GB for 50 accounts.
• Farming Facebook Ads and TikTok Ads accounts — creating and warming up accounts requires different IPs, but the traffic is minimal (scrolling feeds, likes, comments).
• Checking geo-targeting of ads — testing how ads look in different cities and countries. You need IPs from specific regions, but the traffic is just a few pages.
• SMM automation with a small volume of actions — managing client accounts on Instagram, TikTok, VK through ready-made automation services.
• Monitoring positions and competitors — checking Google, Yandex search results from different regions, monitoring competitors' prices on marketplaces.

The main advantage of residential proxies with pay-per-GB payment is the huge pool of IP addresses. You gain access to tens of thousands of real IPs from home users in the countries and cities you need. This is critically important for arbitrage and multi-accounting, where using one IP for multiple accounts leads to instant bans.

Approximate traffic consumption for popular tasks:

Pay-per-IP: when paying for IP addresses is more effective

The pay-per-IP model is justified in situations where you need a stable IP address for a long period and plan to use a large volume of traffic. This is typical for automation tasks, scraping large volumes of data, working with APIs, and any scenarios where IP consistency and high connection speed are important.

When pay-per-IP is more beneficial:

• Intensive scraping of marketplaces — collecting data on thousands of products from Wildberries, Ozon, Yandex.Market. One such scraping can use 10-50 GB of traffic per day.
• Working with social media APIs — automating posting, collecting statistics, managing advertising campaigns through APIs requires a stable IP for authorization.
• Mass account registration — creating hundreds of accounts on marketplaces, social networks, forums. Stable IPs are needed for verification.
• Automating actions on social networks — mass subscriptions, likes, comments through specialized services. High traffic due to content loading.
• Monitoring SEO positions — daily checking of website positions for hundreds of queries in Google and Yandex from different regions.

Especially effective are mobile proxies with the pay-per-IP model for working with platforms that are more lenient towards mobile traffic — Instagram, TikTok, Facebook. Mobile IPs have a high trust score, and one IP can be used for 3-5 accounts simultaneously because mobile network providers use NAT (multiple users behind one IP).

Calculating the profitability of pay-per-IP:

Suppose a mobile proxy costs $100 per month with unlimited traffic. If you use 50 GB of traffic per month, then with the pay-per-GB model ($10 per 1 GB), you would pay $500. The savings would be $400 per month. The break-even point is about 10 GB of traffic per month per IP.

Practical calculations for different tasks

To choose the optimal payment model, you need to forecast your monthly traffic consumption and the required number of IP addresses. Let's consider real cases with calculations.

Case 1: Facebook Ads Arbitrage — 30 accounts

Task: Working with 30 Facebook advertising accounts in Dolphin Anty. Each account is used daily for creating ads, checking statistics, and adjusting bids.

Traffic consumption: One account, when working without video uploads and with image auto-loading disabled, consumes about 5-8 MB per day. Over a month — 150-240 MB per account. For 30 accounts — 4.5-7.2 GB per month.

Cost calculation:

• Pay-per-GB: 7 GB × $10 = $70 per month. Access to thousands of IPs from the required countries.
• Pay-per-IP (mobile): 10 IPs × $100 = $1000 per month (3 accounts per IP). Unlimited traffic.

Conclusion: For this task, pay-per-GB is 14 times more profitable. Residential proxies with payment for traffic are the optimal choice for arbitrage with a small number of accounts.

Case 2: SMM agency — 50 client Instagram accounts

Task: Managing 50 Instagram accounts for clients. Daily posts, stories, responding to comments, monitoring statistics.

Traffic consumption: Instagram consumes more traffic due to high-quality images. One account — 15-25 MB per day, 450-750 MB per month. For 50 accounts — 22.5-37.5 GB per month.

Cost calculation:

• Pay-per-GB: 35 GB × $10 = $350 per month.
• Pay-per-IP (mobile): 15 IPs × $100 = $1500 per month (3-4 accounts per IP).

Conclusion: Pay-per-GB is more profitable, but the difference is smaller (4 times). If traffic exceeds 50 GB, consider a combined approach — some accounts on mobile proxies, some on residential.

Case 3: Scraping Wildberries — 5000 products daily

Task: Daily collection of data on 5000 products on Wildberries — prices, stock, ratings, reviews. Using a ready-made scraper.

Traffic consumption: One product with complete information — about 200-300 KB of data. 5000 products — 1-1.5 GB per day, 30-45 GB per month.

Cost calculation:

• Pay-per-GB: 45 GB × $10 = $450 per month.
• Pay-per-IP (data centers): 5 IPs × $30 = $150 per month (IP rotation to bypass limits).

Conclusion: For intensive scraping, pay-per-IP with data center proxies is 3 times more profitable. Data centers provide high speed, which is critical when processing thousands of requests.

Arbitrage and SMM: which model to choose

For traffic arbitrage and SMM specialists, the choice of payment model depends on the scale of work and the type of platform. Let's consider recommendations for popular scenarios.

Facebook Ads and Google Ads arbitrage

Recommendation: Pay-per-GB with residential proxies.

Justification: Advertising platforms require a unique IP for each account. At the same time, the actual traffic is minimal — you log into the ad cabinet, set up campaigns, check statistics. Residential proxies provide access to millions of real IPs, which is critical for scaling. One account consumes 150-300 MB per month, which at a cost of $10 per GB gives $1.5-3 per account.

Setup in Dolphin Anty: Create a profile for each account, in the proxy settings select a residential proxy with the geo of your target country. Enable IP rotation for each new session or use a sticky session (fixing the IP for 10-30 minutes).

Instagram and TikTok for business

Recommendation: Hybrid approach — mobile proxies (pay-per-IP) for main accounts + residential (pay-per-GB) for additional ones.

Justification: Instagram and TikTok are more lenient towards mobile IPs. If you have 10-15 key client accounts that bring in the main income, place them on mobile proxies — this will provide maximum protection against bans. For the remaining 30-40 accounts, use residential proxies with pay-per-GB.

Calculation: 5 mobile IPs × $100 = $500 (15 accounts, 3 per IP) + 15 GB residential × $10 = $150 (35 accounts). Total $650 for 50 accounts.

Mass subscriptions and automation

Recommendation: Pay-per-IP with mobile proxies.

Justification: Automating actions (subscriptions, likes, comments) generates high traffic due to constant content loading. At the same time, IP stability is needed — frequent IP changes during automation can raise suspicions from the platform. Mobile proxies with unlimited traffic and rotation every few hours are the optimal option.

Scraping and automation: traffic calculation

Scraping and automation are tasks where traffic consumption can vary from a few gigabytes to terabytes per month. Proper traffic calculation is critical for choosing a payment model.

How to calculate traffic for scraping

Step 1: Determine the size of one page. Open DevTools in your browser (F12), go to the Network tab, load the page, and check the total size of the loaded data. Typically, this is 200-800 KB for a product page on a marketplace, 50-150 KB for a search results page.

Step 2: Multiply the page size by the number of requests per day. For example, scraping 1000 products per day × 300 KB = 300 MB per day = 9 GB per month.

Step 3: Add a buffer of 30-50% for repeated requests due to errors, timeouts, data updates. 9 GB × 1.5 = 13.5 GB per month.

Step 4: Compare costs: 13.5 GB × $10 = $135 (pay-per-GB) vs $30-50 for one data center proxy with unlimited traffic (pay-per-IP). If traffic exceeds 5 GB, pay-per-IP is more profitable.

Scraping with API vs HTML scraping

API: If the platform provides an API (e.g., VKontakte, Twitter, YouTube), traffic will be minimal — the API returns only the necessary data in JSON without unnecessary HTML, CSS, images. One request — 5-20 KB. Pay-per-GB is suitable for APIs.

HTML scraping: Scraping full HTML pages (Wildberries, Ozon, Avito) requires loading the entire content of the page. One request — 200-500 KB. With large volumes (thousands of pages per day), pay-per-IP is more profitable.

Traffic optimization when scraping

If you are using pay-per-GB, it is critically important to optimize traffic consumption:

• Disable image loading — most scrapers have an option to load only HTML without images. This reduces traffic by 3-5 times.
• Use caching — save already loaded pages so you don't have to request them again.
• Scrape only the necessary data — if you only need the product price, don't load the entire page, use XPath or CSS selectors for targeted requests.
• Set up compression — make sure your scraper supports gzip compression (header Accept-Encoding: gzip). This reduces traffic by 60-80%.

Hybrid approach: combining models

The most effective strategy for most businesses is to combine both payment models depending on the task. This allows minimizing costs and maximizing flexibility.

Scenario 1: SMM agency with different types of clients

Structure:

• VIP clients (10 accounts): Place on mobile proxies pay-per-IP. Cost: 4 IPs × $100 = $400. Maximum protection against bans, unlimited traffic.
• Standard clients (40 accounts): Use residential proxies pay-per-GB. Consumption: 20 GB × $10 = $200. Access to different geo, traffic optimization.

Total: $600 for 50 accounts instead of $1500 (only mobile) or $350 (only residential with the risk of exceeding traffic limits).

Scenario 2: Arbitrage with testing and scaling

Structure:

• Testing bundles (20 accounts): Residential proxies pay-per-GB. Consumption: 5 GB × $10 = $50. Quick testing of different geo without large investments.
• Working bundles (10 accounts): Mobile proxies pay-per-IP. Cost: 3 IPs × $100 = $300. Stable IPs for accounts that bring profit.

Advantage: You do not spend $100/month on each test account that may not bring profit. Move to mobile proxies only verified bundles.

Scenario 3: E-commerce with scraping and analytics

Structure:

• Daily scraping of competitors (large volume): Data center proxies pay-per-IP. Cost: 5 IPs × $30 = $150. Unlimited traffic, high speed.
• Checking geo-availability (one-time tasks): Residential proxies pay-per-GB. Consumption: 2 GB × $10 = $20. Access to IPs from all regions of Russia.

Total: $170 instead of $500+ when using only residential proxies for all tasks.

Common mistakes when choosing a payment model

A wrong choice of payment model can lead to overpayment by several times or to technical problems. Let's discuss common mistakes.

Mistake 1: Using pay-per-GB for high traffic tasks

Example: You are launching scraping on Wildberries for 10,000 products per day. Traffic consumption — 3 GB per day, 90 GB per month. At a cost of $10 per GB, this is $900 per month. A data center proxy with unlimited traffic costs $30-50.

Solution: Before launching the project, test traffic consumption on a small sample (100-500 requests) and extrapolate for a month. If traffic exceeds 10 GB, consider pay-per-IP.

Mistake 2: Buying pay-per-IP for one-time tasks

Example: You need to check how your site looks from 20 different cities in Russia. You buy 20 mobile IPs for $100 per month = $2000. The task is completed in 1 hour, and the proxies remain idle for the other 29 days.

Solution: For one-time tasks, always use pay-per-GB. 20 checks × 5 MB = 100 MB = $1 instead of $2000.

Mistake 3: Ignoring traffic optimization

Example: You are working with 30 Facebook accounts in an anti-detect browser with image and video auto-loading enabled. Traffic consumption — 50 GB instead of the planned 5 GB. Overpayment — $450.

Solution: In the settings of Dolphin Anty, AdsPower, Multilogin, disable media auto-loading. Install extensions to block ads and trackers. Monitor traffic consumption weekly.

Mistake 4: Using cheap data center proxies for social networks

Example: You buy a data center proxy for $5 to work with Instagram because it has unlimited traffic. A week later, all accounts are blocked — Instagram easily identifies data center IPs and blocks them.

Solution: For social networks (Instagram, TikTok, Facebook), use only residential or mobile proxies. Data centers are suitable only for scraping and tasks where a high trust score IP is not required.

Mistake 5: Incorrect IP rotation settings

Example: You are working with 30 Facebook accounts in an anti-detect browser with incorrect IP rotation settings, leading to frequent bans.

Solution: Ensure that your IP rotation settings are correctly configured to avoid detection and bans.