A First Look at Thordata’s Proxy and Scraping Infrastructure
Upon visiting Thordata’s website, the first thing that strikes me is the sheer breadth of offerings packed into a single platform. Thordata markets itself as a high-quality proxy service for web data scraping, but the dashboard reveals far more: residential proxies (over 100 million IPs across 190+ countries), mobile proxies, static ISP proxies, and datacenter proxies sit alongside a suite of scraping APIs—SERP API, Web Scraper API (with 120+ prebuilt scrapers), Web Unlocker, and a Scraping Browser. The site also lists ready-to-use datasets and, notably, video data solutions tailored for AI model training, including a 6-billion-video dataset from 700 million channels. The layout is clean, with clear product categories and pricing displayed inline. I signed up for the free trial (no credit card required) and quickly tested the residential proxy rotator via a simple cURL command from the docs. The response was fast, the IP correctly masked, and no CAPTCHA blocked the request. For a journalist who routinely scrapes review aggregators, this felt like a significant upgrade over DIY proxy management.
How Thordata Serves AI and Data Teams
Thordata is not just another proxy broker; it’s a full-stack infrastructure play for web data acquisition, especially relevant for AI programming. The platform’s “Data for AI” section explicitly targets LLM and multimodal model training, offering video datasets and a video data scraper that integrates with cloud storage. The Web Unlocker uses intelligent header rotation and browser fingerprinting to bypass anti-bot measures, while the Scraping Browser (powered by stealth browser technology) allows headless script execution with full rendering. During testing, the SERP API returned real-time Google results with structure identical to the organic search page—a boon for anyone building training data for search-related AI models. Under the hood, Thordata doesn’t specify which models power its unlocker or proxy rotation logic, but the 99.9% uptime claim and Chrome extension suggest robust engineering. Integrations include Python, Node.js, and Java code samples in the docs, plus a public API for programmatic control. For developers, this means less time fighting blocks and more time training models.
Pricing and Key Differentiators
Thordata’s pricing is transparent and competitive: residential proxies start at $1.05/GB (with a 50% off promo), mobile proxies at $2.20/GB, static ISP at $0.75/IP, and datacenter proxies at $0.75/IP. Scraping APIs are billed per request: $0.70/1K for SERP API, $0.50/1K for Web Scraper API, $1.00/1K for Web Unlocker, and $2.5/GB for the Scraping Browser. Datasets cost $0.25/1K records. Unlike Oxylabs or Bright Data, which often require minimum spend or annual commitments, Thordata offers a free trial and no credit card required for sign-up. The reseller and affiliate programs (up to 50% commission) are also unique. However, one limitation is the lack of a dedicated AI-specific pricing tier for video datasets—the “Data for AI” section is still priced per record or per GB, which can scale unpredictably for large training runs. Also, the platform doesn’t offer a built-in data labeling or cleaning tool, which some competitors (like Apify or Zyte) provide as add-ons.
Who Should Use Thordata (and Who Shouldn’t)
Thordata is best suited for AI startups, data scientists, and enterprise data teams that need reliable, unblockable proxies and pre-built scrapers to feed AI models with real-world web data. The video datasets and scraper are particularly strong for multimodal model research or video LLM training. Small-scale hobbyists may find the per-GB pricing expensive compared to free solutions like Scrapy with free proxies, but for production-grade scraping at scale, Thordata’s reliability (99.9% uptime) and 190-country coverage justify the cost. Who should look elsewhere? Anyone needing a simple, single-proxy solution for casual use (they have that, but it’s overkill) or teams that require integrated data cleaning/labeling—Thordata currently lacks such features. Overall, Thordata bridges the gap between raw proxy pools and AI data pipelines with remarkable polish. Their focus on video data for LLMs sets them apart in an increasingly crowded market. Visit Thordata at https://thordata.com/ to explore it yourself.
Comments