Browserhub MCP. Run web scraping jobs and manage proxies via conversation.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Browserhub MCP Server automates web scraping and data extraction workflows directly through your AI agent. You don't need to write custom scrapers or manage proxies manually; you just tell your client what data you need from any website, and it runs the job for you.
It handles everything from running one-off scrapes using `direct_scrape` to managing complex jobs with pre-defined blueprints.
What your AI agents can do
Direct scrape
Runs a one-time scrape on any URL when you don't have a pre-written scraper configured.
Get account balance
Checks how many credits you have left in your Browserhub account for scraping jobs.
Get blueprint
Retrieves the full details and settings for a specific, saved scraper blueprint.
Use direct_scrape to pull immediate data from any single URL without requiring a pre-built scraper.
List, retrieve details for (get_scraper, get_blueprint), or list all configured scraping templates to understand your available tools.
Start a job using run_scraper with specific credentials, then track its real-time status and final results via list_scraping_jobs or get_scraping_job.
Verify your account credit using get_account_balance and see which proxy locations are available with list_proxy_locations.
Use list_blueprints to get a complete overview of every saved data extraction pattern you've set up.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Browserhub MCP Server: 10 Tools for Web Data Scraping
These ten tools give your agent full control over the entire scraping process—from running a single URL scrape to managing complex proxy and job queues.
019d7564direct scrape
Runs a one-time scrape on any URL when you don't have a pre-written scraper configured.
019d7564get account balance
Checks how many credits you have left in your Browserhub account for scraping jobs.
019d7564get blueprint
Retrieves the full details and settings for a specific, saved scraper blueprint.
019d7564get scraper
Gets detailed information about an individual configured web scraper.
019d7564get scraping job
Checks the current status and final results of a specific, past scraping job ID.
019d7564list blueprints
Lists all available scraper blueprints, showing which ones you've created.
019d7564list proxy locations
Shows a list of all geographical locations for proxy servers you can use in your scrapes.
019d7564list scrapers
Lists every single scraper configuration that is currently set up on your account.
019d7564list scraping jobs
Retrieves a list of all scraping jobs you've run, along with their basic status and IDs.
019d7564run scraper
Starts an actual data extraction job using one of your configured scrapers on a specific URL.
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 Browserhub, 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
This server lets your agent run web scraping jobs directly from your chat window. You connect it to your AI client—Claude, Cursor, whatever—and forget about managing proxies or writing custom code. Your agent handles everything; you just tell it what data you need off any site.
Need quick info right now? Use direct_scrape. It pulls immediate data from any single URL. You don't have to build a scraper first; you just point it at the page, and it runs the job for you. This is killer for one-off checks.
If you're doing this regularly, you’ll want structure. The system lets you manage complex scraping blueprints. You can check out all your saved templates with list_blueprints, giving you a clean overview of every data extraction pattern you've set up. Wanna see the nuts and bolts of one template? Use get_blueprint to pull its full details and settings.
For managing the scrapers themselves, there are two tools. First, list_scrapers shows you every single scraper configuration currently active on your account. If you need deep specs on just one, use get_scraper. These tools keep track of what you've built so you know exactly what capabilities you're working with.
When it's time to work, you start a job using run_scraper, telling the system which configured scraper and what URL to hit. You can then monitor what’s happening in real-time or check the final results of past runs. To see all your previous attempts and their basic status codes, call list_scraping_jobs.
If you need the deep dive on a single job—the actual results and its final status—use get_scraping_job with the specific job ID.
Don't forget to check your infrastructure. First off, you gotta make sure you got credits; run get_account_balance to see how many scraping credits are left in your Browserhub account. And if rate limits or geography is a thing, use list_proxy_locations to pull the full list of proxy server locations available for your scrapes.
Your agent handles the whole workflow: it checks your balance first, reviews the right blueprint, runs the job using run_scraper, and then reports back on success or failure. You just talk to your client like you're talking to a coworker; it takes care of the API calls behind the scenes.
How Browserhub MCP Works
- 1 First, connect your API key and subscribe to the Browserhub MCP Server.
- 2 Next, prompt your AI agent with a specific web scraping request (e.g., "Scrape this URL using my 'Price Checker' blueprint").
- 3 The server executes the necessary tools (
run_scraperordirect_scrape) and sends the resulting structured data back to your chat interface.
The bottom line is: you tell your AI agent what website content you need, and it runs the scraping job using Browserhub's infrastructure, returning clean, usable data.
Who Is Browserhub MCP For?
Data Scientists who spend too much time writing boilerplate web scrapers. Market Researchers who track competitive pricing across dozens of sites weekly. Developers building agent-based apps that need external web context.
Needs to monitor product availability or price changes across 10 different competitor websites daily. They use the server to run multiple run_scraper jobs in a single chat session.
Requires bulk web data for training machine learning models, but doesn't want to manage proxy lists or job queues manually. They rely on the agent to orchestrate list_scrapers and run_scraper.
Builds applications that need to ingest external web context (like current stock data) without a full backend service. They use the server's tools directly via natural language calls.
What Changes When You Connect
- Zero setup for ad-hoc data: Need a quick scrape? Forget writing code. Use
direct_scrapeto pull structured data from any URL instantly, no blueprints required. - Full job lifecycle control: You don't have to guess if a scrape worked. Use
list_scraping_jobsandget_scraping_jobto track progress in real-time—from initiation (run_scraper) to final data retrieval. - Know your scraper stack: Keep all your scraping logic organized. List everything with
list_scrapersor review the technical specs of a template usingget_blueprint. - Manage infrastructure on demand: Running scrapes costs money and needs good IPs. Check your budget with
get_account_balancebefore you start, and uselist_proxy_locationsto see where your job will run. - Streamlined data retrieval: The best part is that the output isn't just raw text; it's structured data ready for your workflow, pulled directly into your chat context.
Real-World Use Cases
Tracking competitive pricing changes
A market analyst needs to know if a competitor changed their price on three different product pages. Instead of visiting the sites manually, they ask their agent: "Run job for URL A using 'Price Checker' blueprint, and repeat for URLs B and C." The server executes run_scraper three times, returning structured data that highlights only the price changes.
Quickly checking a single article's content
A writer needs to pull key quotes from an article they just found. They don't have a template for it. They ask their agent: "Scrape this URL for the main body text." The server uses direct_scrape, bypassing all blueprints, and hands over clean text immediately.
Auditing job history before billing
A developer wants to confirm if a large data pull was successful or if it failed halfway through. They ask the agent: "List my last 5 jobs." The server calls list_scraping_jobs and they can then check status details using get_scraping_job before verifying their credit with get_account_balance.
Setting up a new data source pipeline
A data scientist wants to use a brand new, complex scraper. They first run list_scrapers to see what's available, then check the detailed setup with get_scraper, and finally initiate the test run using run_scraper.
The Tradeoffs
Treating web scraping like simple text retrieval
Trying to scrape complex, dynamic data (like product tables) by just pointing the agent at the URL. The agent will only get the visible HTML source code, which is messy and unusable.
→
You must use a defined mechanism. Use run_scraper with a specific blueprint, or if it's a quick check, rely on direct_scrape. These tools ensure the data is extracted using real browsers.
Over-relying on manual scraping
Running 20 different web scrapes manually in sequence. This takes forever, and you'll lose track of which job succeeded or failed.
→
Use list_scraping_jobs to queue up multiple tasks, then monitor the batch progress using get_scraping_job. It keeps everything centralized.
Forgetting resource management
Running a massive scrape job without checking if you have enough credits or if your IP address is blocked.
→
Always check get_account_balance first. If the job fails, try listing available IPs with list_proxy_locations to see if switching proxies helps.
When It Fits, When It Doesn't
Use this server when your goal is structured data extraction from websites. You need reliable mechanisms to pull specific fields (like price or name) from a defined URL, not just raw HTML dumps. The strength here is the combination of template management (list_blueprints) and ad-hoc scraping (direct_scrape).
Don't use this if you are performing real-time user interaction testing—for example, clicking through checkout pages or submitting forms that require state changes. This server scrapes data; it doesn't simulate a full browser session. If your task requires knowing how a button feels to click, you need a different kind of testing tool. However, if you just need the text or the price visible on the final page, this is exactly what you need.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Browserhub. 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
Collecting data from websites shouldn't feel like running 50 separate browser tabs.
Today, gathering web data means opening a dozen browser windows. You manually visit site A to find the price; you switch to site B to check availability; then you copy the data into a spreadsheet and repeat that process for sites C through Z. It's slow, tedious, and prone to human error.
With this MCP server, your agent handles it all. You tell your client: "Check the pricing on these five URLs." The system calls `run_scraper` in the background, manages proxy rotation, and returns a clean table of results—no tabs open, no copy-pasting required.
Browserhub MCP Server lets you run data extraction jobs instantly.
You don't have to write Python scripts or manage complex environment variables just to pull a single piece of information. If all you need is the content from one specific URL, the `direct_scrape` tool handles it in seconds, without needing any pre-configured scrapers.
It changes everything. You move from being a manual data collector to an agent who simply directs the flow of information. It's that simple.
Common Questions About Browserhub MCP
How do I use Browserhub MCP Server to check my account balance? +
You call get_account_balance. This tool immediately tells you how many credits remain, helping you budget for large scraping jobs before they start.
Can I scrape a website without having a blueprint? Use direct_scrape. +
Yes. The direct_scrape tool lets your agent run a one-time extraction on any URL, perfect when you just need data fast and don't want to build a full scraper.
What is the difference between list_scrapers and list_blueprints? +
list_scrapers lists every scrapable entity on your account. list_blueprints specifically shows saved, reusable templates—the patterns you've defined for complex data extraction.
How do I know if my scraping job finished successfully? +
You use get_scraping_job. This tool takes the job ID and gives you the current status, telling you if it succeeded or failed and why.
What specific details does `list_proxy_locations` provide? +
It gives you a list of all available proxy locations, including their region and IP type. Your agent uses this data to select the best location for scraping, ensuring your job pulls geo-specific results.
How can I check my job history using `list_scraping_jobs`? +
You can filter the list by status and date range. This lets your agent quickly locate specific jobs—like failed runs or successful reports—without manually sifting through all previous activity.
When I run a job, what should I check first using `get_scraping_job`? +
Check the job status and any associated error logs. The result provides detailed failure reasons or confirms successful data extraction. This lets your agent debug failed runs immediately.
What information does `get_scraper` return for a configuration? +
It returns all metadata for that specific scraper, including its last used URL and required parameters. Use this to confirm the scraper is up-to-date and ready to run with your current data context.
Can I scrape a URL directly without creating a scraper first? +
Yes! Use the direct_scrape tool and provide the URL. It will use Browserhub's real browsers to render the page and extract data based on your parameters instantly.
How do I monitor the progress of a scraping job? +
Simply ask the agent to get_scraping_job and provide the Job ID. It will return the current status (e.g., pending, running, finished) and the result data once ready.
Can I specify a proxy location for my scraping request? +
Yes. When using direct_scrape, you can provide a proxy_country code. Use the list_proxy_locations tool to see the full list of supported countries in your account.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Metaplane
Data observability via Metaplane — track monitors, incidents, and data quality metrics.
RocketLaunch.Live
Track rocket launches worldwide — search upcoming launches, vehicles, companies, locations and missions.
H2O.ai
Manage AI models via H2O.ai — track data frames, monitor machine learning models and training jobs, and audit cloud cluster status directly from any AI agent.
You might also like
Bloomberg Law
Access 200M+ court dockets, case law, and legal news via Bloomberg Law Enterprise Dockets API for comprehensive legal research.
Hyrule Compendium
Access the comprehensive Breath of the Wild and Tears of the Kingdom compendium — search creatures, monsters, materials, and equipment directly from your AI agent.
MessageFlow
Omnichannel communication via MessageFlow — send SMS, WhatsApp, and Email messages.