<- all tokdocs
2026.06.11@chase_ai_

How to Build a Free LinkedIn Scraper With n8n and Google's Custom Search API

AI
Watch on TikTok

View on TikTok ->

Chase walks through a LinkedIn scraper he built inside n8n as a tool for his AI personal assistant, "Jarvis." The scraper pulls up to 1,000 LinkedIn profiles per day at zero cost by using Google's Custom Search API, and it dumps the results straight into Google Sheets. He built the whole thing in about 15 minutes with no code.

What Jarvis Actually Is

Jarvis is a large n8n-based AI agent that acts as Chase's personal assistant. The final frame of the video shows the full workflow: a central "Jarvis" tools agent node connected to a Postgres chat memory store, a Supabase vector store for RAG, email tools, a calendar tool, a calculator, a contacts sheet, an internet tool, and now the LinkedIn scraper. Inputs come through Telegram voice messages, which get transcribed by OpenAI and summarized before output. The architecture is entirely visual and built without writing code.

How the LinkedIn Scraper Works

The scraper workflow visible in the video has five distinct nodes arranged in a pipeline:

  1. Parse User Input - An OpenAI node interprets natural language requests like "give me 100 profiles of software engineers working in Chicago in tech"
  2. Pagination - A code node handles batching so the scraper can work through multiple pages of results
  3. Scrape 100 Profiles - An HTTP Request node hits Google's Custom Search API (the URL visible in the node points to customsearch.googleapis.com)
  4. Filter Data - A code node strips the response down to just the profile URL, name, and title
  5. Upload Results - A Google Sheets node appends the cleaned data, with an "Edit Fields" node for final formatting

The workflow header notes three key features: up to 1,000 profiles per day before hitting pay limits, extraction of job titles, industry, and location, and the ability to adapt the scraper for other sites like Twitter or GitHub.

Why This Matters for Cold Outreach

Paid LinkedIn scraping tools (PhantomBuster, Apify LinkedIn scrapers, etc.) typically cost $50-200/month for similar volume. Chase's approach uses Google's free tier of the Custom Search API, which provides 100 queries per day at no charge. By structuring each query to return 10 results, that gets you to 1,000 profiles daily without spending anything.

The tradeoff is that Google Custom Search results are less detailed than direct LinkedIn scraping. You get the public-facing profile snippet that Google indexes (name, title, URL), not the full profile data that tools with LinkedIn API access or browser automation can pull. For basic prospecting lists, though, the name-title-URL combination is often enough to start outreach.

The n8n Advantage

n8n is an open-source workflow automation platform that competes with Zapier and Make. The key difference for a build like this is that n8n lets you self-host, which means no per-execution pricing. Chase's entire Jarvis agent, including the LinkedIn scraper, the RAG pipeline, and all the connected tools, runs as a single n8n instance. That keeps the ongoing cost at whatever you pay for hosting (or nothing if you run it locally).

Key Takeaways

  • Google Custom Search API gives you 1,000 LinkedIn profile lookups per day for free by combining 100 free queries with 10 results per query
  • n8n's visual workflow builder makes it possible to wire up the scraper, data filtering, and Google Sheets export without writing code
  • The scraper is a tool inside a larger agent, not a standalone workflow, which means Jarvis can decide when to use it based on conversational context
  • The same pattern works for other sites since Google Custom Search can be scoped to any domain, not just LinkedIn
  • Build time was about 15 minutes, which gives a sense of how lightweight these n8n tool integrations are once you have the parent agent set up

Resources

Published June 11, 2026. Writeup generated from a favorited TikTok.