Your Browser Fingerprint Is Clean. Your IP Isn't.

A multi-account operation usually fails at the network layer, not the cookie layer. An anti-detect browser can give every profile its own canvas, fonts, and timezone, and it still cannot change the fact that twelve profiles egress from one datacenter subnet. Thordata exists for that second half of the problem: a residential proxy network of 100M+ IPs across 190+ countries, sold pay-as-you-go from 1 GB with no contract, at a rate that starts at $2.00/GB and steps down to $0.65/GB as volume grows.

Key takeaways

• Fingerprint isolation and IP isolation solve different detection vectors; you need both.
• Residential IPs are simply the addresses real shoppers and real viewers use, which is why they stop "impossible" questions like what does this listing cost in Phoenix.
• Thordata's residential pool covers 190+ countries with city-level targeting, HTTP/HTTPS/SOCKS5 support, and sticky sessions from 30 to 90 minutes.
• Pricing is a usage ladder rather than a committed plan: 1 GB at $2.00/GB, 50 GB at $1.50/GB, 350 GB at $0.80/GB, 5 TB at $0.65/GB.
• Static ISP and datacenter IPs bill per IP ($1.50 standard, $0.75 on promo), and the scraping APIs start around $0.50–$0.80 per 1,000 results with roughly 5,000 free credits to test them.

The ban email rarely mentions cookies

Ask anyone who has run an account matrix for six months how the first suspension happened. It is almost never dramatic. A TikTok seller in Shenzhen gets a "suspicious login activity" notice on an account that had been quiet for weeks. A media buyer finds that four of her eight ad-test profiles now share one review queue. A dropshipper's Amazon Seller Central session starts failing CAPTCHA on every page load.

In each case the browser profiles were fine. What tied them together was the exit: one VPS IP, one VPN endpoint, or one shared office line that every profile happened to touch during the same window. Platforms score network provenance aggressively because it is cheap to compute. A /24 block that has been seen logging into 400 unrelated accounts is a strong signal regardless of how convincing each session's canvas looked.

An anti-detect browser handles device identity. Residential proxies handle network identity. Treating them as interchangeable is the most common and most expensive mistake in this workflow.

What "residential" actually buys you

Residential IPs are addresses assigned by consumer ISPs to households. When a platform looks at traffic from one, the geolocation database says a home in Leeds, the ASN says a broadband provider, and the account behaviour around it says a person. That is the entire value proposition, and it is why Thordata's pool is described in terms of households rather than servers.

The practical parameters that matter when you plug one of these networks into browser profiles:

• Coverage. 190+ countries, with targeting down to city level, so a profile pretending to be Austin behaves like Austin at the network layer too.
• Protocols. HTTP, HTTPS, and SOCKS5, which is what every anti-detect browser's proxy configuration screen expects.
• Session persistence. Sticky sessions from 30 to 90 minutes, long enough to survive a checkout flow or a multi-page ad review without the IP changing mid-session and tripping a risk rule.
• Rotation. On-request rotation per profile, so profile A and profile B never share an exit even when they run on the same machine.

Configure it once in the browser's proxy settings and the profile keeps a stable, local-looking exit for its whole life. Nothing else in your stack has to change — that is deliberate. Most teams are not looking for a new platform; they are looking for an IP source that drops into what they already run.

Four workflows that quietly return wrong answers

The account-safety argument is the obvious one. Less obvious is how much operational data is distorted when the exit is in the wrong place, and this is where teams lose money without noticing.

Price and assortment monitoring. A lot of commerce surfaces adjust to the visitor's region. A datacenter IP in New Jersey reads one price; a household connection in Madrid reads another, sometimes with different stock, shipping estimates, and coupon availability. Monitoring from the wrong exit gives you a clean-looking dataset that describes a market you are not selling into.

Search and marketplace visibility checks. "Where do we rank?" is only answerable from the geography you care about. National-level granularity is often not enough for marketplaces and app stores, which is why city targeting tends to become a hard requirement once the first report gets challenged in a meeting.

Ad verification. Creative rendering, placement, and even whether an ad serves at all can differ by market. Verification from a residential exit is the only way to see what the paying audience sees, which matters most when you are auditing a demand-side partner rather than your own dashboard.

Dataset collection for models and analytics. Teams assembling training or benchmark corpora need geographic and demographic diversity in the source, plus a defensible provenance story. Repeated blocked requests and silent rate-limiting are what usually push these projects toward a managed collection layer instead of hand-rolled fetchers.

The second problem is the invoice

Network identity is a solved category technically. What still goes wrong commercially is procurement: teams either buy a committed plan that assumes usage they do not have yet, or they buy nothing and pay per-request premiums all year.

Thordata's residential pricing is a usage ladder with no minimum commitment. The published steps, captured from the official rate card in mid-2026, look like this:

Two things follow from that structure. First, a two-person team can start on 1 GB and learn whether the network behaves well on their targets before committing anything. Second, the "from $0.65/GB" figure everyone quotes is the 5 TB step, so budget against your actual volume rather than the headline. At 50 GB the effective cost is about $75; at 350 GB, about $280.

For comparison, Oxylabs publishes committed monthly tiers — 5 GB at $30, 20 GB at $100, 125 GB at $500, and 1 TB at $2,500 — which is a different procurement shape rather than a better or worse one. Enterprise-oriented networks such as Bright Data and Decodo generally work the same way: negotiated volume agreements, account management, and rate cards built for organisations that want contracts. If your finance team prefers a purchase order and a fixed annual commitment, that model is legitimate and those vendors are built for it. If you would rather pay for the gigabytes you actually route and keep the option to change your mind, a pay-as-you-go network is simply a better fit.

Beyond residential, the same account covers static ISP and datacenter IPs billed per IP ($1.50 standard, $0.75 during current promotions), and mobile 4G/5G traffic on its own ladder from $5.00/GB at entry down to $2.20/GB at 500 GB. Mixed estates are common: long-lived seller accounts on sticky ISP addresses, high-churn research on residential, app-store checks on mobile.

When you would use an API instead of an IP

Residential IPs are the right tool when the browser or the automation needs to be someone. They are the wrong tool when you just want structured output from a search engine or a marketplace page and do not want to maintain a scraper.

That is what the API line covers. The SERP API runs roughly $0.70–$0.80 per 1,000 results, the Web Scraper API starts around $0.50 per 1,000 results at volume, and the Web Unlocker sits near $1.00–$1.30 per 1,000 requests, with a scraping browser priced on traffic between $5.00 and $2.50 per GB. Each of those ships with around 5,000 free credits, enough to run a real comparison against whatever you use today without talking to a salesperson.

A practical pattern that works well for Chrovia users: keep browser profiles on residential IPs for anything authenticated or session-sensitive, and push unauthenticated collection to the APIs where you are paying per result instead of per gigabyte. Most teams discover their traffic splits roughly that way once they look.

Getting started without a migration project

There is no integration to build. You create an account, generate credentials, choose country or city, pick sticky or rotating, and paste the host, port, user, and password into the browser's proxy field. Rotation is controlled by session duration rather than by API calls, which is usually what anti-detect browsers want.

If you want to test the collection side first, start with the free API credits against three or four of your hardest targets and compare success rates to your current setup over a week. That is a fairer evaluation than any benchmark table, including this one.

A note on responsible use

Proxy networks are infrastructure. Account operations must respect each platform's terms, and data collection must respect applicable law and the site terms that govern it. Nothing here is a licence to evade enforcement, and teams that use residential traffic to hide policy violations tend to discover that network quality is the least of their problems eventually.

Frequently asked questions

Does a residential proxy replace an anti-detect browser? No. They cover different signals. An anti-detect browser controls what the page can learn about your device and browser environment; a residential proxy controls what your network address reveals about where you connect from. Running multiple profiles on one non-residential exit is a common cause of association, and running one profile through a residential IP while leaking a mismatched device fingerprint does not help either.

How much traffic does one account profile actually use? Interactive browsing over residential proxies typically consumes tens to a few hundred megabytes per day per active profile, depending on how much media the target sites load. Automated collection burns through it much faster. Start with a small 1 GB purchase, measure for a week, then size the next month properly.

Is city-level targeting really necessary? For marketplace pricing, app-store rankings, and regional ad verification, usually yes. Country-level exits can sit in a different media market entirely, and a national average is exactly the number that gets challenged in a review meeting.

What is the difference between residential and static ISP proxies? Residential addresses are assigned dynamically to households and rotate on your schedule. Static ISP addresses belong to the same networks but are rented to you for a longer period, so a profile keeps one familiar address, which some platforms reward. ISP addresses are billed per IP here, at $1.50 standard and $0.75 on promotion.

Can I use Thordata alongside my existing vendor rather than replacing it? Plenty of teams run two suppliers deliberately, for redundancy and for regional coverage differences. Pay-as-you-go billing makes a second source cheap to keep warm; a small residential allocation is enough to validate failover behaviour.

Do the free API credits require a sales conversation? No. The scraping, SERP, and unlocker APIs come with around 5,000 credits for evaluation. Residential trials are arranged through support rather than self-serve, so ask when you sign up.

The short version

Fingerprint hygiene gets you a believable device. A residential exit gets you a believable location, and location is also what makes your collected data worth reading. If your current setup forces you to guess between those two, or to sign a volume commitment before you know your own usage, look at the pay-as-you-go option and measure it against your real targets for a week.