Ukraine Proxy: 1.2M+ Residential IPs

Read the terms first. Public pages only, and follow the site's terms:

• Prom.ua: its user agreement (clause 4.4) bars devices, programs or automatic processes that access, copy or track marketplace content, and clause 4.8 bars copying information from the site. Ask Prom.ua for permission, or use your own seller data, before any automated run.
• Rozetka: read its user agreement before you automate.
• Every other shop: check its terms the same way; a proxy changes the IP a site sees, not what its terms allow.

Test on the Easy Trial first. 100 of your own listing URLs at about 0.5 MB each is about 50 MB, under $0.15 of the 1 GB for $2.95. Count how many come back with status 200 and a hryvnia price; that count, and your measured page weight, tell you what the month will cost before you buy it.

What a daily run costs. At about 0.5 MB per listing page (an example size; weigh your own pages), 30 days a month:

The Easy Trial (1 GB for $2.95) covers about 2,000 such pages: enough to measure your page weight and how many pages come back in hryvnias before you buy the month.

1. Confirm the Ukrainian exit IP. The first line prints the IP, the second the country code, which should be UA.

curl -x gate.proxyfield.io:7000 -U "<login>-country-ua:<password>" https://api.ipify.org
curl -x gate.proxyfield.io:7000 -U "<login>-country-ua:<password>" https://ipinfo.io/country

2. Check that the pool is deep enough for your run. Ask for 20 Ukrainian IPs with the plain username (a new IP per request) and count the distinct ones:

n=$(for i in $(seq 20); do
      curl -s -x gate.proxyfield.io:7000 -U "<login>-country-ua:<password>" https://api.ipify.org; echo
    done | sort -u | wc -l)
echo "ua: $n distinct IPs of 20"

The output is one line, ua: <n> distinct IPs of 20. A count at or near 20 means a 1,000-page day is spread over many home connections; a count well below 20 means IPs repeat sooner and you should slow the run. The 20 requests use well under 1 MB of the Easy Trial.

3. Fetch a listing page in Ukrainian, on one sticky IP while you page through it:

curl -x gate.proxyfield.io:7000 -U "<login>-country-ua-session-he9f2qga-ttl-600:<password>" \
  -H "Accept-Language: uk-UA" "https://<listing page URL>"

4. Save each price with its currency label and the time of the check, so a daily history can be compared.

5. Collect politely. Read the site's robots.txt (for example https://prom.ua/robots.txt) and leave out the paths it disallows; keep to 2 or 3 parallel requests per site; when a site answers HTTP 429 (Too Many Requests) or 503 (Service Unavailable), wait, honouring its Retry-After header if it sends one, instead of retrying at once.

The daily job in Python: each page gets a fresh sticky session, the hryvnia check runs before anything is saved, a 429 or 503 makes the job wait (the site's Retry-After, or 60 seconds), and a page that comes back blocked or in the wrong currency is retried on a new Ukrainian IP at most twice, then logged and skipped.

import time, uuid, requests

def fetch_uah(url, tries=3):
    for attempt in range(1, tries + 1):
        sid = uuid.uuid4().hex[:8]            # a new session id = a new Ukrainian IP
        proxy = f"http://<login>-country-ua-session-{sid}-ttl-600:<password>@gate.proxyfield.io:7000"
        try:
            r = requests.get(url, proxies={"http": proxy, "https": proxy}, timeout=30,
                             headers={"Accept-Language": "uk-UA"})
            if r.status_code == 200 and ("₴" in r.text or "грн" in r.text):
                return r.text                 # a page priced in hryvnias: parse and save it
            if r.status_code in (429, 503):   # the site asks you to slow down: wait, do not hammer
                wait = r.headers.get("Retry-After", "60")
                time.sleep(int(wait) if wait.isdigit() else 60)
            print(f"attempt {attempt}: status {r.status_code}, no hryvnia price")
        except requests.RequestException as e:
            print(f"attempt {attempt}: {e}")
        time.sleep(5)
    print("skipped:", url)                    # log it and move on; do not loop
    return None

The same file, fetch_uah.py, then runs the day's list and appends one row per page to a CSV, so each day's prices sit next to the day before:

import csv, datetime, re

PRICE = re.compile(r"(\d[\d\s\u00a0]*)\s?(?:₴|грн)")   # the first hryvnia amount; swap in your own parser

if __name__ == "__main__":
    today = datetime.date.today().isoformat()
    urls = [u.strip() for u in open("urls.txt", encoding="utf-8") if u.strip()]
    with open("prices.csv", "a", newline="", encoding="utf-8") as f:
        out = csv.writer(f)
        for url in urls:
            html = fetch_uah(url)
            m = PRICE.search(html) if html else None
            price = re.sub(r"\s", "", m.group(1)) if m else ""
            out.writerow([today, url, price, "UAH" if price else ""])
            time.sleep(5)                     # a few seconds between pages

Each run adds rows in the form date,url,price,currency, such as 2026-10-28,https://<listing page URL>,1299,UAH; a page that was skipped gets an empty price, so gaps show in the file. Schedule it once a day with cron (here at 06:00 server time):

# crontab -e
0 6 * * * cd /home/you/uah && /usr/bin/python3 fetch_uah.py >> fetch.log 2>&1

Run it over your list once a day, a few seconds between pages and no more than 2 or 3 pages of one site at a time. At 1,000 pages and 5 seconds a page, one run takes a little over 1 h 23 min on a single thread (5,000 s of pauses plus the fetches).

The traffic you do not use stays on your balance. If you would rather receive the pages than run the requests, the Web Scraping API (in development; API documentation) is priced at $0.50 per 1,000 successful results for plain HTTP, or $0.20 from 10,000 results in one purchase: Contact sales. See e-commerce scraping, proxies for e-commerce and price monitoring.