Deploy and Host Wigolo Web Search for AI Agents on Railway

Wigolo is "the web for your AI coding agent": a self-hosted search, fetch, crawl and extraction engine with on-device ranking and embeddings. It exposes everything as a remote MCP server (StreamableHTTP + SSE) and a REST API with an OpenAPI contract, so Claude Code, Cursor, Codex, n8n, LangChain, CrewAI, LlamaIndex or your own services can search and read the web without per-query search-API fees.

About Hosting Wigolo

This template runs the official ghcr.io/knockoutez/wigolo image in HTTP server mode (wigolo serve) on port 3333 with:

a persistent volume at /data for the model cache, browser engine and page cache
a randomly generated WIGOLO_API_TOKEN — Wigolo is fail-closed, so every MCP/REST request must send Authorization: Bearer YOUR_TOKEN (only /health is open, for the Railway healthcheck)
a public HTTPS domain

After deploy:

Copy WIGOLO_API_TOKEN from the service's Variables tab.
MCP clients: add a remote MCP server at https://YOUR-RAILWAY-DOMAIN/mcp with header Authorization: Bearer YOUR_TOKEN (legacy clients: /sse).
REST: POST https://YOUR-RAILWAY-DOMAIN/v1/search with a JSON body like {"query": "..."}; the full contract is at /openapi.json.

The first requests are slower while models and the browser engine download into the volume.

Common Use Cases

Give coding agents (Claude Code, Cursor, Codex, OpenCode…) real web search and page reading via MCP
Web research nodes in n8n / Make / custom automations through the REST API
Retrieval for RAG pipelines (search, find-similar, crawl and extract to clean content)
A shared, team-wide search backend instead of paying per query for a hosted search API

Dependencies for Wigolo Hosting

Docker image ghcr.io/knockoutez/wigolo
Persistent volume (≈1–2 GB for models, browser and cache)

Deployment Dependencies

Source: https://github.com/KnockOutEZ/wigolo
Self-hosting guide: https://github.com/KnockOutEZ/wigolo/blob/main/docs/self-hosting.md
REST API reference: https://github.com/KnockOutEZ/wigolo/blob/main/docs/rest-api.md

Implementation Details

Optional: add GEMINI_API_KEY (free tier is enough) to enable LLM-assisted features described in the Wigolo docs. Some challenge-protected sites score IP reputation, so datacenter IPs may be blocked where a home connection is not; see the self-hosting guide for the opt-in proxy setting.

Why Deploy Wigolo on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Wigolo on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.