Oxylabs MCP. Systematically scrape data from Google, Amazon, and custom sites.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Oxylabs connects your AI agent to a full suite of web scraping tools. This server lets you systematically pull structured data from complex websites like Amazon and Google.
You can scrape specific SERPs (Google, Bing, Yandex), deep product listings, or any arbitrary URL—even if it requires JavaScript rendering.
It’s built for comparative market analysis and large-scale data extraction.
What your AI agents can do
Scrape amazon product
Extracts deep e-commerce details for a specific Amazon ASIN, including variations and product specs.
Scrape amazon search
Performs structured data extraction by looping over multiple listing results from an Amazon search query.
Scrape bing serp
Calculates and extracts structured mapping data directly from Bing Search Engine Results Pages.
Pull structured data from Google, Bing, and Yandex search results pages (SERPs) using dedicated parsers.
Retrieve complex data points like variations, pricing history, reviews, and stock status for specific Amazon or Google Shopping products.
Scrape raw HTML from any given URL using the universal scraper, even when client-side JavaScript rendering is required.
Send highly customized JSON payloads to control proxy types, locations, and specific scraping parameters.
Pull URLs, dimensions, and hosting origins directly from Google Image Search results.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Oxylabs MCP Server: 10 Tools for Data Extraction
This server gives you ten specialized tools to pull everything from search engine results pages (SERPs) to deep Amazon product listings.
019d75edscrape amazon product
Extracts deep e-commerce details for a specific Amazon ASIN, including variations and product specs.
019d75edscrape amazon search
Performs structured data extraction by looping over multiple listing results from an Amazon search query.
019d75edscrape bing serp
Calculates and extracts structured mapping data directly from Bing Search Engine Results Pages.
019d75edscrape custom payload
Sends entirely arbitrary, highly customized JSON payloads to control proxy types and advanced scraping settings.
019d75edscrape google images
Parses specific layout blocks from Google Image Search, providing image URLs and technical metadata.
019d75edscrape google news
Exports time-indexed news articles by pulling structured data directly from the Google News format.
019d75edscrape google serp
Extracts structured data blocks from general Google Search results pages (SERPs).
019d75edscrape google shopping
Gathers e-commerce structures by hitting Google Shopping search engines to monitor product listings.
019d75edscrape universal
Scrapes raw HTML from any website endpoint, safely avoiding proxy blocks and supporting JavaScript rendering.
019d75edscrape yandex serp
Performs structural analysis and indexing of results specifically from the Yandex search engine.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Oxylabs, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
You connect your AI agent to a full suite of web scraping tools with this server. It lets you systematically pull structured data from complex websites like Amazon or Google. You can scrape specific SERPs (Google, Bing, Yandex), deep product listings, or any arbitrary URL—even if the site needs JavaScript rendering.
This setup is built for comparative market analysis and pulling large amounts of reliable data.
For general web scraping, use scrape_universal to pull raw HTML from virtually any website endpoint; it handles client-side JavaScript rendering so you don't get blocked by simple scrapers. You can also send highly customized JSON payloads using scrape_custom_payload, giving your agent full control over proxy types, locations, and specific scraping parameters.
When you need to analyze search engine results pages (SERPs), dedicated parsers give you structured data blocks instead of just messy text. If you're checking Google searches, the scrape_google_serp tool pulls structured data from general SERPs. For specialized e-commerce monitoring, use scrape_google_shopping to gather product structures directly from Google Shopping search engines.
When comparing results across major platforms, your agent can run scrape_bing_serp to calculate and extract structured mapping data straight out of Bing Search Engine Results Pages. If Yandex is what you're tracking, the scrape_yandex_serp tool performs structural analysis and indexing specifically for that search engine.
For image metadata collection, run scrape_google_images to parse specific layout blocks from Google Image Search results; this gives you URLs, dimensions, and technical hosting origins. If news articles are your target, scrape_google_news exports time-indexed content by pulling structured data directly from the Google News format.
Deep e-commerce profiling requires more than a simple search—you need specifics. To gather complex data points like variations, pricing history, reviews, and stock status for an Amazon product, you hit scrape_amazon_product using its specific ASIN. If your job is to track multiple listings from one query on Amazon, you'll run scrape_amazon_search; this performs structured data extraction by looping over many results from a single Amazon search query.
When monitoring the deep details of product listings across e-commerce sites, these tools let you go way beyond basic titles.
Your agent can pull structured mapping data directly from Google Search Engine Results Pages (SERPs) using scrape_google_serp. If you need to see what's available on Amazon search results in a repeatable fashion, use scrape_amazon_search again. To monitor the full scope of product listings visible through Google Shopping searches, deploy scrape_google_shopping.
When your research involves parsing specific layout blocks from Google Image Search, remember that scrape_google_images handles pulling those URLs and metadata.
When you're building a data pipeline, think about how these tools stack up. You can use the general-purpose power of scrape_universal to pull raw HTML from an unknown source, then feed that output into structured parsers like scrape_google_serp or scrape_bing_serp for cleaning and analysis. The ability to send tailored instructions via scrape_custom_payload means you control the proxy types and advanced scraping settings your agent uses every time it runs a query.
You're not limited to just search engines; if you encounter an arbitrary URL that needs scraping, scrape_universal is your go-to. It provides raw HTML extraction while safely avoiding common proxy blocks, even when the site relies on client-side JavaScript rendering. The sheer breadth of these specialized tools means whether you need deep product details from Amazon, structured news articles from Google News, or just a basic dump of all the content from an obscure website, you've got the tool for it.
You can check specific e-commerce structures using scrape_google_shopping to monitor how competitors list their products across Google’s platform, and you get comprehensive coverage with dedicated SERP parsers for Google (scrape_google_serp), Bing (scrape_bing_serp), and Yandex (scrape_yandex_serp). This setup gives your agent the power to act like a full-time data analyst, pulling structured insights from any corner of the web.
How Oxylabs MCP Works
- 1 Subscribe to the server and input your Oxylabs username and password.
- 2 From your AI client (Claude, Cursor, etc.), call a specific tool function with the target URL or query parameters.
- 3 The agent runs the scrape, receiving structured data (JSON/HTML) that you can then process further.
The bottom line is: You define what to scrape and where; the server handles the complex task of rendering JavaScript and bypassing site protections.
Who Is Oxylabs MCP For?
This is for data analysts, SEO teams, e-commerce strategists, and market researchers. If your job requires comparing pricing or tracking rankings across multiple live websites—especially Amazon or Google—you need this. It saves you from managing proxies and fighting anti-bot measures.
Uses scrape_google_serp and scrape_yandex_serp to track keyword ranking changes across multiple countries without buying proxy lists.
Runs scrape_amazon_product repeatedly on competitor ASINs, pulling structured data points like price variance and feature differences for comparative reports.
Employs scrape_universal to gather raw content from niche websites or uses specialized tools like scrape_google_shopping to monitor competitor pricing across regional markets.
What Changes When You Connect
- Cross-Platform Comparison: Instead of building separate scrapers for each search engine, you use dedicated tools like
scrape_google_serp,scrape_bing_serp, andscrape_yandex_serpto pull comparative data from Google, Bing, and Yandex in a single workflow. - Deep E-commerce Insight: You get more than just the title. Use
scrape_amazon_productto capture variations, stock states, and seller data for an ASIN—essential when monitoring competitor inventory across regional domains. - Handling Modern Websites: If your target site is a single-page app (SPA) that requires client-side JS execution, the
scrape_universaltool handles this safely. You just setjs_render=true, and it gets the raw HTML you need. - Targeted Search Results: Need to track brand mentions? Use
scrape_google_newsfor time-indexed articles orscrape_google_imagesto pull image metadata, rather than relying on general text scraping. - Advanced Control Flow: You aren't limited by the toolset. The
scrape_custom_payloadlets you send fully customized JSON payloads—great if you need specific proxy configurations or deeply nested parameters. - Broad Coverage: Combine search and shopping data: Use
scrape_google_shoppingto see what Google thinks is a product, then usescrape_amazon_searchon that query to compare the actual e-commerce listings.
Real-World Use Cases
Monitoring Amazon competitor pricing across regions
An analyst needs to track how a competitor's flagship product (ASIN X) is priced in both the US and UK markets. They use scrape_amazon_product twice—once for amazon.com and once for amazon.co.uk—to build a single, comparative price matrix.
Gathering comprehensive search result context
A content strategist needs to understand the full scope of competition for 'best noise cancelling headphones'. They run scrape_google_serp and then follow up with scrape_google_shopping on the same query. This provides both organic ranking data and actionable product listing information.
Scraping a complex, modern marketing site
A researcher needs to scrape an enterprise client's custom dashboard (which runs entirely in JavaScript). They can't use simple HTTP requests. Instead, they run scrape_universal with js_render=true, ensuring the agent gets the full rendered content for analysis.
Analyzing global search trends
A global marketing team wants to see how their keyword ranks across different countries. They use scrape_google_serp for US results, then scrape_yandex_serp for the Russian market, and finally scrape_bing_serp for European analysis—all structured data extraction.
The Tradeoffs
Trying to scrape a modern SPA with basic HTTP requests
The agent just hits the main URL and gets empty results because it ignores JavaScript. It fails to pull the actual product cards or dynamic content.
→
Don't use general scraping endpoints without rendering capability. Instead, call scrape_universal and explicitly set the parameter js_render=true. This forces the agent to wait until the full client-side page is built before extracting the HTML.
Scraping product details using a generic search scraper
The user runs scrape_google_serp and tries to extract product data from the results. The output is messy, incomplete, or lacks crucial fields like Prime status.
→
For reliable e-commerce data, use dedicated tools. If you need a single product, run scrape_amazon_product. For searching listings, use scrape_amazon_search.
Ignoring location or proxy needs
The agent fails because the target site blocks requests from default locations. The resulting data is inaccurate or empty.
→
If you need advanced geo-targeting or specific proxy controls, don't guess. Use scrape_custom_payload to send a fully customized JSON schema payload detailing exactly what location and proxy types you require.
When It Fits, When It Doesn't
Use this server if your primary goal is comparative data analysis across multiple large platforms (Amazon vs. Google) or deep, structured extraction from complex web pages. You need to know what the product page says, not just that it exists.
Don't use this if you just need simple text scraping of a static blog post—a simpler HTTP library might suffice. More critically, don't use it if your process requires real-time data streaming (e.g., live chat logs). This server is for structured batch extraction. If you are building an agent that needs to constantly verify inventory in milliseconds, the complexity and rate limits of multiple scraping tools might introduce too much friction; consider a dedicated database API instead.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Oxylabs. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually pulling data from competitor websites takes hours.
Right now? You open Amazon.com, check the main product page. Then you switch tabs to Google Shopping to see if they list it there too. Next, you have to manually go through a search result on Bing to catch any related articles. It's clicking between 4 or 5 different services, copying URLs, and re-running requests just to build one coherent report.
With Oxylabs, your agent handles the whole mess. You tell it: 'Pull product data for ASIN X from Amazon, and then pull all related results from Google Shopping.' The tools—like `scrape_amazon_product` and `scrape_google_shopping`—run these tasks sequentially, giving you a single JSON output that compares all three points without you ever leaving the chat window.
The Oxylabs MCP Server: structured data from any source.
Before this server, getting robust search results meant picking a tool and accepting its limitations. If you needed Google SERPs *and* Amazon listings, you had to run two separate scripts that didn't talk to each other. You were always dealing with siloed data sets.
Now, your agent coordinates the calls. It runs `scrape_google_serp` for context and then uses a targeted tool like `scrape_amazon_search` for specific product listings. The result is a unified knowledge graph: you get both broad market context *and* deep product detail in one go.
Common Questions About Oxylabs MCP
Can I scrape any website with the scrape_universal tool? +
Yes, scrape_universal is designed to pull raw HTML from any endpoint. If the site uses JavaScript for rendering (like an SPA), you must use js_render=true in your payload.
How do I compare Amazon and Google Shopping listings? +
You run two separate tools: scrape_amazon_product for the deep details on one ASIN, and then use scrape_google_shopping to pull the general listing data. The agent will give you both sets of structured results.
Which tool is best for tracking keyword rankings? +
Use scrape_google_serp if you need standard Google SERPs, or use scrape_yandex_serp if your market focus is the CIS region. Both provide structured ranking data.
Do I need to manually configure proxy types? +
No, but you can if you need more control. If you require a specific setup (like residential proxies in a certain country), use scrape_custom_payload to send the precise JSON instructions.
How do I set up my credentials for advanced tools like `scrape_custom_payload`? +
You must first subscribe to the server and input your Oxylabs username and password. Once authenticated through Vinkius, your AI client routes all requests using these stored API credentials.
If a website changes its structure, how reliable is `scrape_google_serp`? +
The tool maintains specialized parsers for SERPs. While minor structural shifts are handled automatically, major platform overhauls might require you to adjust the underlying schema payload.
What format does `scrape_amazon_product` return the extracted data in? +
It delivers highly structured JSON objects. The tool doesn't give you raw HTML; instead, it populates specific keys with fields like ASINs, prices, and review counts for immediate agent use.
What are the limitations if I run `scrape_google_shopping` at high volume? +
Oxylabs manages rate limits to ensure stable access. For large-scale, sustained data gathering, you should implement a controlled batching system or check their dedicated API guidelines.
Can the scraper handle JavaScript-heavy websites like React or Vue apps? +
Yes. Enable js_render=true on the universal scraper and Oxylabs runs a headless Chromium instance that executes all client-side JavaScript before extracting the HTML. This handles SPAs, lazy-loaded content, and dynamically rendered components.
How does it handle anti-bot protection? +
Oxylabs manages a global residential and datacenter proxy network with automatic IP rotation, browser fingerprinting, and CAPTCHA solving. The API handles all proxy management transparently — you just send the URL and receive structured data back. No proxy configuration needed on your end.
Can I scrape Amazon in different countries — like amazon.de or amazon.co.jp? +
Yes. Both Amazon search and product scrapers accept a domain parameter (com, de, co.jp, co.uk, fr, it, es, etc.). The parser adapts to the regional layout and returns localized pricing, availability, and seller data specific to that marketplace.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
ANOVA Calculator Engine
Run exact One-Way ANOVA tests to compare means across multiple groups local. Get CPU-guaranteed F-scores and p-values, not LLM guesses.
Groq
Run large language models at unprecedented speed with custom LPU hardware that delivers real-time AI inference at massive scale.
Nodereal
High-performance blockchain infrastructure for Aptos and EVM networks — query accounts, blocks, and simulate transactions.
You might also like
Optum Eligibility
Verify patient active healthcare coverage, extract deductibles, and run real-time UHG demographics.
Blastscan (Blast Network Explorer)
Explore the Blast Network—check balances, track transactions, and inspect smart contracts directly from your AI agent.
EventTemple
Fill your venue calendar with event bookings, catering orders, and banquet management tools made for hospitality sales teams.