Firecrawl on Railway
Turn websites into markdown or JSON for LLM apps: self-hosted Firecrawl with an API-key gateway on Railway.
Firecrawl is an open-source API that scrapes and crawls websites and returns clean markdown, HTML, screenshots or structured JSON. It is built for LLM applications such as RAG pipelines and AI agents that need web content in a usable form.
Developers call it through its SDKs, its MCP server or integrations such as n8n and LangChain. It also offers a hosted cloud service; self-hosting keeps requests and data on your own infrastructure.
Firecrawl is licensed under AGPL-3.0. Check the licence before building a network service on top of a modified copy.
What Firecrawl can do
- Scrape and crawl endpoints returning markdown, HTML or JSON
- Headless Chromium rendering for JavaScript-heavy pages
- Queued crawling with separate workers
- Structured extraction with an OpenAI-compatible model (optional)
- SDKs and MCP server for agents
- BullMQ queue dashboard
What people use it for
- Converting documentation sites you may crawl into markdown for RAG
- Giving AI agents a scrape and search tool
- Extracting structured data from product or listing pages
- Monitoring your own sites for content changes
Videos
Running Firecrawl on Railway
The Railway template deploys the self-hosted stack as separate services: a Caddy gateway that only forwards requests with your generated API key, the Firecrawl API, an orchestration worker, scrape executors, a Playwright (Chromium) service, a Postgres job queue, Redis and RabbitMQ. All databases stay on the private network. Several always-running services plus Chromium make memory the main cost driver. JSON extraction needs an OpenAI-compatible API key that you add yourself.
Firecrawl templates on Railway
Tracklayer templates are listed first. Other Firecrawl templates follow, ordered by total deployments on Railway. How this works.
Deployment numbers as of 2026-11-20.
Questions about Firecrawl
How do I call the API?
Copy FIRECRAWL_API_KEY from the gateway service's variables and send it as a Bearer token to the gateway's public URL, for example to /v2/scrape. The SDKs accept the gateway URL as the API URL.
Is the instance open to anyone?
Not with this template. Self-hosted Firecrawl has no cloud authentication backend, so the template puts a gateway in front that rejects requests without your key.
Do I need an OpenAI key?
Only for JSON extraction and the /extract endpoint. Set OPENAI_API_KEY on the firecrawl-api service. Plain markdown scraping works without it.
Can I crawl any website?
You are responsible for crawling only sites you are permitted to, and for following Railway's Fair Use Policy.
What is the licence?
AGPL-3.0. Running it for your own use is fine; read the licence if you modify it and offer it as a service to others.
Guides
Similar apps
Tracklayer maintains the templates marked as such; Firecrawl is a trademark of its owners and this site is not affiliated with them.
