Proxies for Web Scraping
Real 4G/5G mobile IPs for public data collectionMobile proxies for web scraping that rotate per request across real carrier IPs, hold a sticky session when a crawl needs one, and bill by the gigabyte so a lean crawl stays cheap.
From $3.99 per GB, or a dedicated device from $49 a month.
Why a pay-per-GB pool fits web scraping
Crawls are bursty and parallel: many short sessions from many exits, not one identity held for weeks
Per-request rotation across a whole country pool is the default, sticky sessions are one username parameter away
Billing by traffic rewards lean crawls: strip media, cache what does not change, pay for the data you needed
Recommended configuration for web scraping
The settings you would set in the dashboard, with the reasoning for this workload.
- Rotation
- Per request for catalog and search crawls
- Switch to a sticky session when a target paginates behind a cookie
- Sticky session
- Add _s_<id> to the flex username to hold one exit
- Add _ttl_ to cap it, otherwise it expires after an hour idle
- Concurrency
- Scale parallel sessions with the pool
- Keep per-domain request rates at human pace and back off on any 429
- Protocol
- HTTP/HTTPS or SOCKS5 with the credentials from the dashboard
- Works with Scrapy, Playwright, Puppeteer and plain fetch
AI agents
An agent can run this Web Scraping setup
An AI agent can buy the plan, pull the gateway credentials, rotate the IP and watch usage through VoidMob's open-source MCP server, so it can provision the Web Scraping setup above without a human in the dashboard. Works with Claude, Codex and any MCP client.
What goes wrong with proxies for web scraping
Symptom, cause, fix. Each one traces back to a setting on this page.
The bill grows faster than the dataset
Full page renders pull images, fonts and video the parser never reads, and every megabyte is billed
Block media at the browser or proxy layer and request JSON endpoints where they exist. Most catalog and pricing crawls transfer a fraction of the page weight once media is off.
Paginated flows reset to page one mid-crawl
The exit IP rotated inside a flow that the site binds to a cookie, so the session looked hijacked
Use a sticky session for the length of a paginated or logged-in flow, then release it. Rotate between flows, not inside them.
Check with IP CheckerOne domain starts returning 429 or empty pages
Requests hit a single carrier IP at datacenter speed; carrier IPs are shared by real subscribers, so sites rate-limit them conservatively
Spread requests across exits and keep per-IP cadence close to what a phone would do. Treat a 429 as a signal to slow down, not to retry harder.
Prices and availability in the data do not match what users in the market see
The crawl exited from one country while the data is regional
Run the same crawl from each target country's pool and store the exit country with every row. Country, city and ASN targeting are username parameters on the pool.
Check with Location Test
Why mobile IPs work for web scraping
Public data collection runs cleanest from IPs that look like ordinary phone subscribers. VoidMob routes each request through a real 4G/5G device behind carrier-grade NAT, rotates the exit per request or holds it for a session, and bills by the gigabyte, so a crawl costs what it transfers.
Web scraping lives or dies on how a target classifies the source IP. Datacenter ranges are catalogued and throttled on sight; residential ranges are better but increasingly fingerprinted. Mobile carrier ranges sit behind carrier-grade NAT, where one public IP fronts thousands of real subscribers, so a site that blocks a mobile IP blocks its own users. That economic fact is why a request from a real 4G/5G device is treated like a phone on a commute rather than a script, and why mobile proxies have become the default for public-data crawls that need to keep running week after week. The CGNAT explainer covers the mechanics.
VoidMob's pool is built from physical devices on consumer carrier plans, not virtual slices of a datacenter uplink. Each request exits with that carrier's ASN and DNS, and the flex username lets you pick per-request rotation or a sticky session so a paginated flow stays on one identity. Billing is per gigabyte with no per-request fee, which rewards lean crawls. The calculator on this page turns pages per day and page weight into a rough monthly figure so you can pick a tier before you start.
Collect public data, respect robots.txt and each site's terms, and keep per-domain rates modest. The network makes the vantage point honest; what you collect, how fast, and whether it contains personal data are your calls, and the Playwright and Puppeteer guides show the connection strings that keep a headless browser on one exit.
What it costs
Pay-per-GB shared pools, or dedicated devices billed per day, week or month.
Pay-per-GB
HTTP(S) and SOCKS5. Rotate per request or hold a session.
- 1 GB$3.99 ($3.99/GB)
- 10 GB$35 ($3.50/GB)
- 25 GB$81.25 ($3.25/GB)
- 50 GB$149.50 ($2.99/GB)
- 100 GB$250 ($2.50/GB)
- Above 100 GBCustom volume plans
Dedicated devices
HTTP(S), SOCKS5 and OpenVPN on every device. VLESS/Xray on premium tiers.
United Statesfrom $129/mo
United Kingdomfrom $97/mo
Germanyfrom $119/mo
Francefrom $79/mo
Brasilfrom $99/mo
Indonesiafrom $72/mo
Italyfrom $127/mo
Spainfrom $119/mo
Canadafrom $159/mo
Polandfrom $59/mo
Thailandfrom $59/mo
Australiafrom $119/mo
Singaporefrom $119/mo
Ukrainefrom $49/mo
Romaniafrom $49/mo
Austriafrom $79/mo
Irelandfrom $59/mo
Netherlandsfrom $139/mo
Portugalfrom $89/mo
New Zealandfrom $129/mo
Georgiafrom $109/mo
Lithuaniafrom $59/mo
Moldovafrom $55/mo
Prices updated
How much data does web scraping use?
Pre-filled with a typical setup for this use case. Adjust the inputs to match yours.
Rough estimate, not a billing forecast. Real usage varies by target, media mix and caching.
About 0.2 MB per page without media, over 30 days.
Prices updated
Run Web Scraping on real mobile IPs
Pools from $3.99/GB, or a dedicated device for 1 to 30 days.
Frequently Asked Questions
Common questions about mobile proxies for web scraping.