WyndPath, the scraping API that finds its way

What WyndPath does

All of scraping, one API.

Proxies, browser, anti-bot and geolocation handled for you, behind a single API.

Optimized per target

WyndPath knows the best setup for each site (route, proxy, rendering) and applies it on its own. You send nothing but a URL.

Pay-per-success

No credit charged on a failure. A max_cost cap per request rules out surprises.

Geolocation

Choose the country the request comes from and see the web like a local visitor. Several countries available.

Rotation & reliability

Automatic IP rotation, response validation and a retry when a request is blocked. A high success rate, with no intervention.

Traced routes

Every request is logged (route, IP, latency, credits). Per-target stats learn the best route over time.

The heart of the product

Get the page, without a second thought

JavaScript challenges, cookies, sessions, page rendering: WyndPath gets the page where a plain proxy fails. That's the difference between a page you got and a request that got blocked.

How it works

Three steps, one URL.

1

You send

A key, a target URL. Nothing else to know.

2

WyndPath routes

It detects the target, picks the route and the IP, then gets the page.

3

You receive

The raw page and the route headers. Billed on success.

request

# one key, one URL. WyndPath takes care of the rest.
GET https://api.wyndpath.com/v1/
    ?api_key=wk_9f3c…              # your key
    &url=https%3A%2F%2Fexample.com   # the target
    &country=fr                      # optional

X-WyndPath-Route: browser          # route chosen
X-WyndPath-Credits: 10
X-WyndPath-Status: 200 OK          # page rendered

AI agents

Give your AI access to reliable web data.

Your assistant tries its normal web access first. When the page comes back empty, blocked or fully rendered in JavaScript, it finds WyndPath among its tools and goes and gets the page itself, without being told to.

REST API

For applications and scripts.

Read the docs →

Proxy

For browsers and scraping frameworks.

Read the docs →

New

MCP

For AI assistants and autonomous agents.

Learn more →

mcp.json

// one URL, WyndPath shows up as a tool
{
  "mcpServers": {
    "wyndpath": { "type": "http", "url": "https://mcp.wyndpath.com/mcp" }
  }
}

Reading

Understand the playing field.

Clear articles on scraping and anti-bot systems, backed by figures.

Figures

How much of web traffic is generated by bots?

One visit in two is no longer human. The story in numbers.

Read →

Technical

How anti-bot systems work

The layers that decide whether your request gets through or gets blocked.

Read →

Guide

Residential, datacenter and mobile proxies

The type of IP changes everything. Differences, costs, use cases.

Read →

All articles →

Frequently asked questions

What you need to know.

What is WyndPath?

A scraping API: you send a URL, WyndPath picks the route, handles the proxies, the JavaScript rendering, the sessions and the cookies, then returns the page. One HTTP call, hosted in Europe.

How is WyndPath billed?

In credits, on success only: a failed call consumes no credits. A simple site costs 1 credit, a protected one costs more, and a per-request cap avoids surprises. You get 1,000 credits free to start.

Is web scraping legal?

Collecting public data is legal in France, but regulated: GDPR for personal data, database rights, respect for the robots.txt file. WyndPath applies rate limiting and respects the directives published by sites.

Do I have to manage proxies myself?

No. WyndPath handles datacenter, residential and mobile proxies, along with geolocation. You can optionally choose the country, and the rest is automatic.

Does WyndPath work on sites that use Cloudflare or DataDome?

Yes. Where a plain proxy fails, WyndPath uses a real browser, runs the JavaScript and presents a coherent fingerprint, within the rules published by the sites.

Can I choose the country the request comes from?

Yes, via the country parameter. Several countries are available, which lets you see a site like a local visitor.

Which languages can I use it with?

Any of them: it is a standard HTTP API. Ready-to-use examples exist in Python, Node.js, PHP, Go, Ruby, Java, C# and R.

Practical guides