Residential Proxies - Real IPs, Zero Blocks

Your connection string

One package gives you one login. Country, city and session lifetime ride along in the login itself, so you never open a ticket to add a location.

Endpoint

proxy.sotaproxy.com:10000

login[_c_{COUNTRY}][_city_{City-Name}][_s_{sessionId}][_ttl_{30s|15m|1h}]:password@proxy.sotaproxy.com:10000

curl -x login:password@proxy.sotaproxy.com:10000 https://api.ipify.org

curl -x login_c_US:password@proxy.sotaproxy.com:10000 https://api.ipify.org

curl -x login_c_US_city_New-York_s_100_ttl_15m:password@proxy.sotaproxy.com:10000 https://api.ipify.org

City names take a hyphen instead of a space. Country codes are two letters, ISO 3166-1 alpha-2. Suffixes stack in the order shown above.

Rotating and sticky sessions

Leave off _s_ and every request goes out through a different address. That suits scraping where each page stands alone: search results, product listings, price checks. Add _s_{id} and the address stays put for the lifetime you set in _ttl_.

The session id is yours to choose. Run 200 accounts in parallel by giving each one its own id, _s_1 through _s_200, and each keeps a separate address.

Setup in an antidetect browser

Each profile takes its own session id, which keeps its address stable while the profile is open. In Afina, AdsPower, Dolphin, Octo or Undetectable, open the proxy settings of a profile and fill in:

Change the number after _s_ for every profile. Two profiles sharing an id share an address, which is the link you are trying to avoid.

Setup in code

proxies = {
    "http": "http://login_c_DE:password@proxy.sotaproxy.com:10000",
    "https": "http://login_c_DE:password@proxy.sotaproxy.com:10000",
}
r = requests.get("https://example.com", proxies=proxies, timeout=30)

browser = playwright.chromium.launch(proxy={
    "server": "http://proxy.sotaproxy.com:10000",
    "username": "login_c_GB_s_5_ttl_15m",
    "password": "password",
})

All integrations: Python, Scrapy, Selenium, Playwright and more

Countries and cities

The pool covers 220+ countries. City targeting works in the larger markets, and these are cities our customers run traffic through right now:

Coverage in a given city moves with the devices online at that moment. If _city_ returns a 400, the pool for it is empty right now: drop to country level and retry later.

Price and how traffic is counted

Residential traffic costs $1.00 per gigabyte. No monthly minimum, no subscription, and traffic stays on the package while the package is active. Metering counts both directions: the request you send and the response you get back, headers included.

• Skip images and fonts in your scraper. On a product page they carry most of the weight and none of the data you came for.
• Ask for gzip. Set Accept-Encoding: gzip and an HTML page arrives three to four times smaller.

Errors you will actually hit

Limits

• No cap on concurrent connections and no cap on session count.
• The 30s, 15m and 1h lifetimes are ceilings, not guarantees. A residential device can drop off at any moment and your session ends with it, so handle reconnects.
• Mail ports and a handful of other restricted ports stay closed by default. Our upstream opens them after identity verification, so write to support if your task needs one.

When residential is the wrong choice

Residential traffic is metered per gigabyte, so bulk jobs get expensive. Three cases where another type fits better:

API and AI agents

Everything on this page works through the REST API at https://api.sotaproxy.com/api/v1. We also run an MCP server, so Claude and other agents can check the balance, buy traffic and pull the connection string without a browser.