Decodo vs. Thordata: A Developer’s Side-by-Side Look at Scraping APIs (No Marketing Fluff)

Both Decodo (formerly Smartproxy) and Thordata sell scraping APIs to developers, and both are good at what they do. The decision comes down to three questions: How much do you hate per-GB pricing for structured data work? Do you need SERP data? And does your volume justify result-based pricing over subscription tiers? This article compares the published numbers and gives you a decision framework — not a verdict.

I’ve spent enough time in scraping documentation to know that most “X vs Y” articles are written by people who have never shipped a pipeline. So here’s the deal: no fake benchmarks, no “winner.” Just the published pricing, the API surfaces, and a framework for choosing.

The Products, Plainly

Decodo is the rebranded Smartproxy — a consumer-friendly proxy and scraping platform. Its Scraping API is subscription-tiered ($19/$49/$99 per month + VAT), with a free starter plan, 100+ scraping templates, and a REST endpoint. Its residential proxies run $3.75/GB at 3 GB, down to $2.75/GB at 100 GB, with pay-as-you-go wallet pricing at $8.50/GB. Its Site Unblocker starts at $0.95 per 1,000 requests.

Thordata is a web data platform built around result-based pricing. Its Web Scraper API has 120+ pre-built scrapers (Amazon, eBay, LinkedIn, YouTube, Zillow, Booking.com, and more) with CAPTCHA solving and JS rendering included, priced around $1.00 per 1,000 results at entry volume and about $0.50 per 1,000 at scale. Its SERP API delivers structured Google/Bing results from ~$0.70 per 1,000 responses. Its residential proxies sit on a published volume slider: $2.00/GB at 1 GB down to $0.65/GB at 5,000 GB.

Side by side:

Right away you can see why this isn’t a “which is better” question. The two companies package the same underlying capability differently, and the packaging matches different teams.

Decision Point 1: Subscription vs. Per-Result

Decodo’s subscription model is genuinely attractive below a certain volume. If you’re collecting a few thousand pages per month and want predictable billing, a $19/month tier with a free starter plan is a low-risk entry — and the free plan means you can prototype without a credit card conversation.

The math crosses over fast, though. At 100,000 results per month, Decodo’s top self-serve tier is $99/month. The same volume on Thordata’s per-result pricing at the entry rate is roughly $100 — nearly identical. At 500,000 results per month, subscription tiers force you into multiple plans or sales conversations, while per-result pricing just keeps scaling: 500K results at the volume rate is ~$250–500/month depending on your tier position, with no ceiling to bump into.

Rule of thumb: prototyping and low volume favor subscriptions; production volumes above ~100K results/month favor per-result pricing.

Decision Point 2: SERP Data Changes Everything

If your workload includes search results — rank tracking, SERP feature analysis, AI search visibility — this is the decisive difference. Decodo’s Fast Search API is sales-led with custom pricing. Thordata’s SERP API is self-serve, published, and cheap: ~$0.70 per 1,000 structured responses at volume, with real-time Google and Bing coverage and automatic CAPTCHA handling.

For a concrete example, the same request through Thordata’s API:

from thordata import Thordata

client = Thordata(api_key="YOUR_API_KEY")

result = client.serp(
    query="best project management software 2026",
    search_engine="google",
    location="United States",
    device="desktop",
    output_format="json",
)

# Structured fields — no HTML parsing, ever
for item in result.organic[:5]:
    print(item["position"], item["title"], item["url"])

# Or a scraper-API job for a marketplace target
job = client.scrape(
    scraper="amazon_product",
    query="B08N5WRWNW",
    domain="com",
    output_format="json",
)
print(job.result["price"], job.result["seller"])

And if what you actually want is monitoring rather than one-off queries — scheduled crawls, position history, alerts when ranks move — Thordata’s SERP monitoring solution packages the collection layer into a managed workflow. That’s a category Decodo currently fills through its sales team rather than a self-serve product.

Decision Point 3: The Long-Tail Proxy Story

Every scraping API user eventually hits a target that no pre-built scraper covers, and then the underlying proxy pool becomes the product. Both companies sell raw proxies; the published floors differ:

• Decodo residential: $2.75/GB at the 100 GB tier; $3.75/GB at 3 GB
• Thordata residential: $2.00/GB at 1 GB; $0.73/GB at 1 TB; $0.65/GB at 5,000 GB

At small volumes the prices are close. The divergence is at scale: Thordata’s slider keeps falling to $0.65/GB with no negotiation, which matters when raw-proxy workloads (long-tail sites, custom parsers) grow alongside your API usage. Both support city-level geo-targeting; Thordata adds ASN-level targeting and sticky sessions up to 90 minutes on a network it sizes at 100M+ residential IPs across 190+ countries.

Also worth knowing: Thordata’s Scraping Browser ($2.50/GB) works with Puppeteer, Playwright, Selenium, and Scrapy, and integrates with anti-detect browsers like AdsPower and BitBrowser — relevant if your workflow mixes API calls with scripted browser automation.

FAQ

Which one should a solo developer or student pick? Decodo’s free starter plan is the friendliest possible on-ramp — no cost, low commitment. Thordata’s free trial credits for its scraper and SERP APIs are also zero-cost to evaluate. Honestly: try both, they’re free to test.

Which has better coverage of major e-commerce sites? Both cover the major marketplaces through pre-built scrapers/templates. Thordata’s catalog (120+) is slightly larger and includes LinkedIn, YouTube, and Crunchbase alongside retail targets. Check your specific must-have targets against both catalogs before deciding — this is the one comparison that’s genuinely workload-specific.

I need datasets, not APIs. Does either sell them? Thordata does — pre-built datasets from ~$0.25 per 1,000 records across 100+ domains, custom dataset scoping, and a video dataset built for multimodal AI training. Decodo’s catalog doesn’t currently position datasets as a core product.

What about reliability and support? Both publish uptime claims in the 99.9% range. The practical difference is support motion: Decodo is self-serve with account-based support; Thordata is similarly self-serve, with support via support@thordata.com and a developer docs portal. Neither is an enterprise SLA product at these price points — if you need one, you’re buying enterprise contracts from any vendor. For rank-tracking workloads specifically, Thordata’s managed SERP monitoring adds scheduled collection and alerting on top of the raw API, which narrows the operations gap further.

Can I use both? Yes, and many teams do — Decodo for a stable subscription-billed workload, Thordata for per-result jobs and SERP data. Multi-provider failover is a best practice in production scraping anyway; using two vendors per design is cheaper than using one by accident.

The Framework, Condensed

Pick Decodo if: you’re prototyping, your volume is under ~100K results/month, you value subscription predictability, and SERP data isn’t central to your product.

Pick Thordata if: your volume is above ~100K results/month, you want per-result pricing that scales without tier ceilings, SERP data is part of your workload, or you also need datasets for analytics and AI training.

Still unsure? Price your actual workload on both models — take your monthly request count and multiply it through both pricing structures, then run each API against your top five target domains with the free trials. And if SERP monitoring is the core of your use case, start with the SERP monitoring solution — it’s the fastest way to see structured search data flowing through your own stack, and the same trial credits cover the scraper APIs for the rest of your evaluation. You can explore the full continuous SERP data crawling capability alongside the free-tier options before committing a single dollar.