Introducing Wayfern

Web automation rarely fails because clicking a button is difficult. It fails because running browsers reliably at scale requires much more than launching Chromium: consistent fingerprints, proxy routing, session lifecycle management, live search extraction, and an API surface that automation frameworks can use without custom browser patches.

Wayfern brings those pieces together as browser infrastructure for developers. You connect with familiar tools, control a real browser over the standard Chrome DevTools Protocol, and pay for the resources you actually use.

Bring your existing browser client

Wayfern browser sessions expose CDP over WebSocket. Playwright can connect with chromium.connectOverCDP(), Puppeteer can connect with puppeteer.connect(), and other CDP-compatible clients can use the same endpoint.

That means an application does not need a Wayfern-specific automation language. Existing navigation, extraction, screenshot, and interaction code can continue to use the browser library it already knows.

Each connection creates a session on demand. Location parameters can route a session through a residential exit when a workflow needs a particular country, region, city, or ISP, while sessions without location parameters use the default datacenter exit.

Search without maintaining result-page parsers

Search pages change frequently and differ across providers. The Wayfern Search API handles the browser session and returns one normalized response across Google, Bing, and DuckDuckGo.

Results include normalized titles, URLs, snippets, and ranking information. Providers that expose related searches can also return those suggestions, so an application can expand a query or build a broader research plan without parsing provider-specific markup.

The Search API is useful on its own, but it also works as a discovery step before a browser workflow. An application can find relevant pages through live search and then open selected results in a Wayfern session for deeper interaction.

Give AI agents live web tools through MCP

Paid accounts can expose Wayfern search to compatible AI agents through the Model Context Protocol. The MCP server uses authenticated Streamable HTTP and the same API-token model as the rest of Wayfern.

This makes live search available to agents without embedding Wayfern-specific HTTP requests into every agent framework. Connect the MCP endpoint, provide a scoped token, and the agent can call the advertised search tool using structured inputs and outputs.

The MCP guide includes endpoint details and client configuration. The API reference documents the underlying HTTP interface for applications that prefer direct integration.

One credit balance across the platform

Browser sessions, live search, and captcha solving draw from the same credit balance. Monthly plans include credits, additional credit packs do not expire, and optional residential bandwidth is metered separately only when location routing is used.

The goal is a small, predictable surface: use CDP when you need full browser control, use the Search API when you need normalized results, and use MCP when an AI agent should discover and call the search capability itself.

Start with the developer docs, review pricing, or connect an agent using the MCP setup guide.

Frequently asked questions

Which browser automation clients work with Wayfern?

Wayfern exposes the standard Chrome DevTools Protocol, so it works with Playwright, Puppeteer, and other CDP-compatible clients.

Which search engines does the Search API support?

The Search API supports Google, Bing, and DuckDuckGo and returns normalized organic results plus related searches when the provider exposes them.

Can AI agents call Wayfern directly?

Paid Wayfern accounts can connect compatible AI agents through the authenticated Streamable HTTP MCP endpoint using an API token.