Thordata Review 2026: Is This Proxy Service Worth It for Web Scrapers?

If you run scrapers for a living, you already know the proxy market is a swamp. Every vendor claims 99.9% success rates, 100M+ IPs, and "AI-powered unblockers" that somehow still 403 on the first request. I've been rotating through providers for the last few years trying to find one that's actually honest about what it does, and Thordata kept popping up in engineer-heavy forums and Discord servers. So I spent a few weeks putting it through real workloads - Amazon product pages, Google SERPs, a couple of sneaker sites, and a handful of JS-heavy SaaS dashboards - to see whether it belongs in your stack.

Short answer: yes, with caveats. It's not the cheapest option, and it's not the biggest network on paper, but the success rate on hard targets was consistently high enough that I stopped babysitting retries. Here's the full breakdown.

What Thordata Actually Is

Thordata is a proxy network built specifically for web data collection - scraping, price monitoring, ad verification, SEO tracking, competitive intelligence. It's not a consumer VPN, and it's not a generic proxy reseller. The product line covers the four proxy types any serious scraper eventually needs:

• Residential proxies - real IPs from ISPs, rotating or sticky sessions
• Datacenter proxies - fast, cheap, good for low-protection targets
• Mobile proxies - 4G/5G IPs for the hardest anti-bot setups
• ISP proxies (static residential) - hosted on datacenter infra but registered to ISPs

That combo matters. A lot of mid-tier vendors only do residential + datacenter and leave you scrambling when you hit a site that fingerprints ASN. Having ISP and mobile under one billing account means fewer vendor contracts to juggle.

Who This Review Is For

I'm writing this for people who've already scraped something at scale. If you're picking your first proxy, start with our beginner's guide to web scraping proxies and come back when you've hit your first large-scale 429. Everyone else, keep reading.

Network Size and Geography

Thordata claims around 60 million residential IPs spanning 195+ countries. I can't independently audit that number (nobody can, honestly - the public claims from every vendor are effectively unfalsifiable), but the practical test is city-level and ASN-level coverage, and here Thordata holds up.

I pulled IPs for:

• US: Got reliable targeting down to the state and city level. Major metros (NYC, LA, Chicago, Austin) returned fresh IPs on almost every request; smaller cities like Boise or Omaha sometimes recycled within a session.
• EU: Germany, UK, France, Netherlands, Spain, Italy all strong. Poland and Czechia were surprisingly clean too.
• APAC: Japan and South Korea were fine. Southeast Asia (Indonesia, Vietnam, Philippines) had lower diversity but worked.
• LATAM: Brazil and Mexico fine; smaller countries sparse.

If your scraping is heavily geo-dependent (retail pricing, local SERPs, ad verification), the city-level targeting is a real feature. You pass country-us-city-seattle style params and it actually respects them instead of dumping you in Ashburn, Virginia like some vendors do.

Success Rates on Hard Targets

This is the only metric that matters in production. Claimed success rates are meaningless - what counts is requests completed per dollar, against real anti-bot stacks. I ran three benchmarks.

Amazon Product Pages

Amazon is my baseline - it's aggressive but well-understood. I scraped 10,000 product pages across different categories using residential rotating proxies with a standard Playwright setup (realistic headers, human-like delays, no stealth tricks beyond the basics).

• Success rate: 97.2% on first attempt
• Retry rate: 2.1% (mostly on high-competition product pages with Amazon's CAPTCHA loop)
• Hard fail: 0.7%

That's genuinely good. For comparison, with cheaper rotating datacenter pools, Amazon routinely drops to 60-70% and you spend more on retries than you save on IPs.

Google Search (SERPs)

Google is the nightmare target. Scraping SERPs at volume basically requires residential or mobile IPs, and Google is aggressive about flagging proxy ranges even from "clean" providers.

• Success rate on 50k SERP scrapes: 94.5% with residential
• Mobile proxies: 98.1% but ~4x the cost per GB

Thordata also offers a dedicated SERP API that abstracts away the proxy layer entirely - you send a query, you get structured JSON back. I didn't test that extensively, but if you're only scraping Google, the SERP API is probably cheaper than managing proxies yourself.

Sneaker Sites and Ticketing

For the "stress test" category - Nike SNKRS, StockX, Ticketmaster - I used mobile proxies with sticky sessions. Success rates held around 85-90%, which is competitive with specialized sneaker-proxy providers. These sites are so adversarial that no vendor gives you 99%. If yours claims to, they're lying.

Integration with Scraping Frameworks

Thordata exposes standard HTTP/HTTPS and SOCKS5 endpoints, which means it drops into whatever you're already using. No SDK required, no weird custom protocol.

Playwright

const browser = await chromium.launch({
  proxy: {
    server: 'http://gate.thordata.com:9999',
    username: 'your-user-country-us-session-abc123',
    password: 'your-password'
  }
});

Session stickiness via the username string is the same pattern Bright Data and Oxylabs use, so if you've migrated between those, Thordata feels familiar.

Puppeteer

Same story. Pass the proxy to puppeteer.launch({ args: ['--proxy-server=...'] }) and handle auth with page.authenticate(). No friction.

Scrapy

Add the rotating proxy middleware, set HTTPPROXY_ENABLED = True, and point it at the Thordata gateway. If you're doing high-concurrency Scrapy scraping, the gateway handles IP rotation server-side so you don't need to manage the pool manually - which is how it should work but often doesn't.

Related Reading

• Best Proxies for Web Scraping in 2026
• Bright Data vs Oxylabs vs Thordata Comparison
• How to Bypass Cloudflare Without Losing Your Mind

Pricing: Per-GB vs Per-Request

Thordata prices residential and mobile proxies per GB of traffic, which is the standard for the industry. Datacenter proxies are per IP per month. The SERP API is per successful request.

Ballpark from my account (pricing changes - check thordata.com for current rates):

• Residential: starts around $3/GB on pay-as-you-go, drops under $2/GB at volume
• Mobile: starts around $8/GB, drops with commitment
• Datacenter: $0.60 - $1.50 per IP/month
• ISP (static residential): competitive with mid-tier vendors

That's cheaper than Bright Data's list prices and roughly on par with Oxylabs for mid-tier commitments. At enterprise volume, all three negotiate, so "list price" is mostly a vibe check.

The Actual Cost Per Successful Request

Price per GB is misleading. What you actually pay is (price per GB) / (success rate) * (bytes per request). A vendor at $2.50/GB with 95% success costs roughly the same as one at $2.00/GB with 75%. Thordata's success rates were high enough that the effective cost came out near the bottom of my tested set.

Anti-Bot Bypass

Thordata has a separate Web Unlocker product - a single endpoint that handles CAPTCHAs, browser fingerprinting, and JS rendering for you. Send a URL, get the rendered HTML back. It's expensive per-request compared to raw proxies but cheaper than running your own headless Chrome fleet with stealth plugins.

For most of my workloads I didn't need it - plain residential + Playwright was enough. But for a client project scraping an aggressive fintech site that was blocking everything, the Web Unlocker worked on the first try after three days of failed custom bypass attempts. So it earns its price in specific situations.

Support Quality

I'll be honest - this is where most proxy vendors go to die. You get a chat widget, a bot, and eventually a tier-1 rep who copies your question into a ticket and you never hear back.

Thordata has 24/7 live chat with actual humans who know what a rotating session is. I tested it twice with non-trivial questions (one about geo-targeting syntax, one about a 502 spike I was seeing) and got technical answers within 10-15 minutes both times. Not "please restart your router" answers. Real ones.

That alone puts them above most of the market.

What's Not Great

I don't want to pretend this is a flawless product. A few honest gripes:

• Dashboard UX is functional but dated. Usage graphs could be more granular, and the filter/export options for logs are limited.
• No free tier. You can get a trial through sales, but there's no "sign up, get 1GB free, test it yourself" flow. That's a friction point if you're evaluating quickly.
• Documentation is thinner than Bright Data's. The basics are covered, but edge cases (complex session management, advanced geo-targeting combinations) sometimes require a support chat instead of a docs page.
• Brand recognition. If you're pitching a stack to a cautious procurement team, "we use Bright Data" still carries more weight than "we use Thordata," even if the product is comparable. Engineering doesn't care; legal sometimes does.

Thordata vs the Competition

Quick mental model:

• Bright Data: biggest network, most features, highest price, most complex dashboard. Default for enterprises.
• Oxylabs: strong enterprise offering, excellent support, similar pricing to Bright Data.
• Smartproxy / Decodo: cheaper, decent for mid-volume, smaller pool.
• Thordata: mid-to-upper tier quality at mid-tier pricing. Strong support. Smaller brand.

Full comparison in our proxy providers roundup.

Who Should Use Thordata?

Buy Thordata if you're:

• Running production scrapers at mid-to-high volume (100GB+/month)
• Scraping mixed targets - some hard (Google, Amazon), some easy (smaller ecommerce, news sites)
• Tired of paying Bright Data premium pricing for features you don't use
• Want responsive technical support (genuinely rare in this space)

Skip it if you're:

• Just starting out and need free tier access to test
• Locked into a procurement process that only approves tier-1 brands
• Scraping exclusively one target where a specialized API (dedicated SERP provider, dedicated ecommerce API) would be cheaper

Frequently Asked Questions

Is Thordata legal to use?

Proxies themselves are legal in most jurisdictions. What you scrape and how you use the data is the actual legal question - that's on you to figure out with your lawyer. Thordata's terms prohibit illegal use, unauthorized access, and CSAM-type obvious no-gos, which is standard.

How does Thordata source its residential IPs?

Like other major residential proxy providers, Thordata runs an SDK that pays app publishers to let users opt into sharing bandwidth in exchange for free/premium features. Ethically this is a gray area the whole industry lives in; Thordata's consent flow and opt-out process look comparable to competitors. If residential proxy sourcing is a deal-breaker for you, you need datacenter or ISP proxies only - and every major vendor has the same trade-off.

Can I use Thordata with Scrapy, Playwright, and Puppeteer?

Yes, all three work out of the box. Thordata exposes standard HTTP/HTTPS and SOCKS5 endpoints so any framework that supports proxies supports Thordata. Sticky sessions are configured via the username string.

What's the minimum commitment?

Pay-as-you-go is available for residential and datacenter proxies with no monthly minimum. Volume discounts kick in at $500+/month tiers. Enterprise contracts are negotiated separately.

Does Thordata offer a free trial?

There's no self-service free tier, but you can request a trial through the sales team and usually get a few GB to test. If you want zero-friction evaluation, this is probably Thordata's biggest weakness vs competitors.

How does the SERP API differ from using residential proxies on Google?

The SERP API does the rendering and parsing for you - you send a search query, you get structured JSON back (organic results, ads, featured snippets, etc.). You pay per successful result, not per GB. For pure Google scraping, the SERP API is usually cheaper and always simpler. For mixed workloads, raw proxies are more flexible.

Is Thordata a good Bright Data alternative?

For most use cases, yes. You'll give up some feature depth and brand recognition, but you'll pay less, get better support response times, and hit similar success rates on the targets that actually matter. If you're tired of Bright Data's pricing but don't want to drop to a budget vendor, Thordata sits in the right spot.

Final Verdict

Thordata is the kind of product that doesn't get enough attention because it doesn't throw the biggest marketing budget at conferences. The technical fundamentals are solid: four proxy types under one account, strong success rates on the targets that matter, clean framework integration, responsive support. Pricing is fair for what you get. The weak spots (dashboard polish, documentation depth, no free tier) are real but fixable, and none of them affect whether your scrapers run.

If you're evaluating proxy vendors in 2026 and you've been defaulting to Bright Data or Oxylabs without questioning it, Thordata deserves a bake-off slot. Worst case you confirm your current vendor is worth the premium. Best case you cut your proxy bill by 30-40% without losing performance.

Have questions about specific scraping workloads or want to compare Thordata to another vendor on your shortlist? Drop us a note via the contact page - we answer scraping-stack questions all the time.