Guide
Proxycurl alternatives after the shutdown
Proxycurl closed on 4 July 2025, after LinkedIn sued it. If your product read LinkedIn profiles through it, here is what happened, the three ways to replace it, and where Lempile fits and where it doesn't.
By Lempile · Updated
What happened to Proxycurl
According to its founder, Steven Goh, LinkedIn filed a lawsuit against Proxycurl in January 2025. Proxycurl chose not to fight it and, on 4 July 2025, announced it was shutting down to comply with its legal settlement with LinkedIn, after helping existing customers move off it.
The team went on to build NinjaPear. In a post from 29 April 2026, the founder describes it as the API platform he built after Proxycurl, built around different data, and says it does not scrape LinkedIn. So it isn't a drop-in replacement for LinkedIn profile data.
Three ways to replace it
| Approach | What you get | The trade-off |
|---|---|---|
| Licensed or public-source data | Records about people and companies, without any LinkedIn account | Not the live LinkedIn profile; check how fresh the data is and where it comes from |
| Your users' own LinkedIn accounts, through an API | The live profile, posts, company pages and search, as that account sees them | Each user connects an account; reads are paced per account |
| Your own scraper | Whatever you build | The kind of collection LinkedIn sued Proxycurl over, plus the upkeep |
Where Lempile fits
Lempile is the second approach. Your users connect their LinkedIn account through a hosted link, and your code reads LinkedIn through that account:
- a profile with its sections, or a profile looked up by its URL (
GET /api/v1/users/{identifier},GET /api/v1/users?provider_url=); - company pages (
GET /api/v1/linkedin/company/{identifier}); - LinkedIn's classic search for people, posts, jobs and companies (
POST /api/v1/linkedin/search); - a person's or a company's posts, each with its exact date (
GET /api/v1/users/{identifier}/posts).
If you also used Proxycurl for work emails, Lempile's Email Finder finds them from LinkedIn profile URLs with its own resolver, in jobs you can poll or download as CSV.
Where Lempile doesn't fit
- No account, no read. Every read goes through a LinkedIn account its owner connected. Lempile is not a database.
- Human pace, not bulk. Each account makes at most 1,500 LinkedIn calls a day, one at a time; a profile takes three, and its sections one to five more. Enriching leads as they arrive works; pulling millions of profiles doesn't. See the limits.
- Classic LinkedIn only. Sales Navigator and Recruiter aren't supported.
Which one to pick
If you need company and people records at volume and don't need the live profile, look at licensed data providers. If your product acts on LinkedIn for its users (reading a lead's profile before a message, following their posts, sending the message), reading through the user's own account gives you the live data and the actions in one API. Try the search and profile routes with a free key.