Buying Guides 5 min read

How Much Proxy Bandwidth Do I Need? Measure It, Then Buy It

Stop guessing GB. Measure what one page costs through a proxy, multiply, add a buffer, and buy the right plan size.

Most residential and mobile proxies are billed per GB, so how much proxy bandwidth do I need is really a budget question. Guess too low and your job stops halfway with a 402 error; guess too high and you've prepaid for traffic you won't use. The fix is to measure what one request actually costs, multiply by your volume and add a buffer. Here's how to do each step.

What Counts Toward Proxy Bandwidth

Bandwidth is the data that passes through the proxy in both directions: your request headers and body going out, and the response coming back, including every image, script, font and video chunk if you're using a browser. Response data is almost always the big part. A plain HTTP client fetching HTML downloads only the HTML; a browser downloads everything the page references.

Step 1: Measure One Page

For an HTTP client, measure the response size directly. This Python snippet fetches a page through the proxy and prints the decompressed body size, which is an upper bound on the transfer since most sites compress responses:

import requests

proxy = 'http://USERNAME:[email protected]:8080'
proxies = {'http': proxy, 'https': proxy}

urls = [
    'https://example.com/',
    'https://example.com/page-2',
]

total = 0
for url in urls:
    r = requests.get(url, proxies=proxies, timeout=30)
    size = len(r.content)
    total += size
    print(url, r.status_code, round(size / 1024, 1), 'KB')

print('average KB per page:', round(total / len(urls) / 1024, 1))

For a browser, open DevTools, go to the Network tab, tick Disable cache and reload. The bottom bar shows the transferred total for the page. Load three or four typical pages and average them.

Step 2: Apply the Formula

GB needed = pages x average MB per page x (1 + retry rate) / 1024 x 1.2 buffer

The retry rate is the share of requests you'll repeat because of timeouts, blocks or CAPTCHAs. For a tolerant site use 0.05; for a strict one start at 0.2 and measure. The 1.2 buffer covers redirects, larger-than-average pages and the occasional test run.

Worked Examples

  • HTML scraping: 50,000 product pages at 150 KB each with 10 percent retries. 50,000 x 0.146 MB x 1.1 / 1024 x 1.2 = about 9.4 GB.
  • Search result pages: 300 keywords x 3 countries x 30 days at 250 KB each, 5 percent retries. 27,000 pages, about 8.1 GB a month.
  • Headless browser with images blocked: 10,000 pages at 600 KB each, 15 percent retries. About 7.9 GB.
  • Headless browser with everything loaded: the same 10,000 pages at 3 MB each. About 40 GB. Same job, five times the bandwidth.
  • Managing five social accounts in a browser for an hour a day: highly variable, often 100 to 300 MB per account per day with media. Measure a real day before buying a month.

Step 3: Cut Bandwidth Before You Buy More

The cheapest GB is the one you don't use. In rough order of impact:

  1. Use an HTTP client instead of a browser when the data is in the HTML or an API response. This alone can cut usage by 80 percent or more.
  2. In a browser, block images, media and fonts. In Playwright: page.route('**/*', r => ['image','media','font'].includes(r.request().resourceType()) ? r.abort() : r.continue()).
  3. Send Accept-Encoding: gzip, br so responses arrive compressed. Most HTTP libraries do this by default; check yours.
  4. Call the site's JSON endpoints if they're public and allowed, instead of rendering full pages.
  5. Cache pages you've already fetched and don't re-download unchanged ones.
  6. Fix retry loops. A bug that retries a 403 forever can burn a plan overnight. Cap retries and back off.
  7. Route static or lenient requests through cheaper datacenter bandwidth and keep residential for the strict pages.

More ideas for keeping costs low are in avoid blocks when web scraping, since fewer blocks also means fewer retries.

When Unlimited Bandwidth Makes More Sense

If you're moving a lot of data in a short window, compare per-GB pricing with an hourly plan. NodeProxies unlimited residential costs $3.90 per hour with no GB cap. At the entry residential rate of $0.65/GB the break-even is 6 GB per hour; at the 1000 GB+ rate of $0.35/GB it's about 11 GB per hour. If you'll consistently push more than that, unlimited is cheaper. See unlimited bandwidth residential proxies for more.

Buying Bandwidth on NodeProxies

Once you have a number, buy close to it. NodeProxies plans start at 100 MB, so you can run the measurement step for a few cents, and you can buy exact GB amounts after that. Residential runs from $0.65/GB down to $0.35/GB on 1000 GB+ plans, mobile from $1.10/GB and datacenter from $0.42/GB, all on the pricing page. There's no subscription, and leftover account balance never expires.

FAQ

How much bandwidth does 1 GB of proxies give me?

About 7,000 pages at 150 KB each for plain HTML, or around 340 full pages at 3 MB each in a browser. Measure your own pages for an accurate figure.

Do failed requests use proxy bandwidth?

Any data transferred counts, including error pages and CAPTCHA pages. That's why the retry rate belongs in your estimate.

Does HTTPS use more proxy bandwidth?

Slightly, because of the TLS handshake, but the difference is small compared with page content. It's not worth giving up encryption to save it.

What happens when my proxy bandwidth runs out?

On NodeProxies the gateway returns a 402 status. Buy more bandwidth and requests work again without changing your credentials setup.

NodeProxies

Ready to get started?

90M+ residential, mobile & datacenter proxies. Datacenter from $0.30/GB, residential from $0.42/GB. No contracts.

Create free account