How to extract LinkedIn profile data with Python (legally)

Before any code: LinkedIn is the most aggressively defended scraping target on the public web. Their legal team has filed and won multiple cases against scraper operators, including the long-running hiQ Labs saga that — despite an early hiQ win — eventually settled with hiQ shutting down. If you're doing this commercially, you need a lawyer, not a blog post. With that out of the way, here are the technical options.

1. Official LinkedIn APIs

LinkedIn offers a few API products, all gated:

• Sign In with LinkedIn — basic profile fields for the authenticated user only.
• Marketing Developer Platform — campaign and analytics data, not profile scraping.
• Talent Solutions / Recruiter — full profile data for paying enterprise customers under specific contracts.
• Sales Navigator — lead data, also paid, also restricted.

If you qualify for any of these, use them. They're the only fully ToS-compliant route.

2. The unofficial linkedin-api package

There's a popular Python package on PyPI called linkedin-api that wraps LinkedIn's internal Voyager API using a logged-in session cookie. Functionally it works. Practically:

• It violates LinkedIn's Terms of Service.
• Your account will eventually be flagged or banned.
• Voyager endpoints change without notice, so the library breaks regularly.
• Don't use it from your real account.

If you're going to test it, use a throwaway account and accept that it will not last.

3. Headless browser with Selenium or Playwright

Same ToS issues as above, plus you have to handle login, two-factor, and infinite scroll. LinkedIn's bot detection on its login flow is sophisticated. Don't expect this to be a weekend project.

4. Third-party data providers

Companies like Proxycurl, Phantombuster, and Bright Data sell LinkedIn data through their own infrastructure. They wear the legal and operational risk; you pay per record. For most engineering teams that need LinkedIn data, this is the right tradeoff.

5. Manual export + extraction

If you have a small set of profiles you've legitimately accessed (with permission), you can save the page as PDF or take a screenshot and run it through ExtractFox's LinkedIn extractor or image data extractor. This works well for recruiter shortlists or sales prospect research where the volume is small and human-driven.

Fields you can typically get

• Name, headline, location
• Current and past positions with dates
• Education
• Skills and endorsements
• About / summary text
• Connection count and mutual connections

Practical recommendation for engineers

Default to the official APIs if you qualify. If you don't, buy from a provider — the all-in cost is lower than building and maintaining a scraper, and you offload the legal exposure. Build your own scraping infrastructure only if LinkedIn data is the literal core of your product and you have legal counsel on retainer.