Leaving Proxycurl — ScrapeField

	ScrapeField	Proxycurl (2025)
Status	Operating	Shut down 4 July 2025
What the complaint alleged	n/a	Fake accounts, and access to non-public data
Data accessed	Public and logged-out only	Alleged to include non-public
Surfaces	4 — LinkedIn is one of them, capped at a minority of revenue	LinkedIn only
If one supplier disappears	Another serves the same endpoint, same response shape	The company was the supplier
Credentials	We never take yours, for any reason	n/a
Retention	Cached for a stated TTL, then deleted. No stored profile corpus	Held a large profile database

Checked 20 September 2026 against each vendor’s own pricing page. Rechecked quarterly — tell us if something here is out of date and we will fix it.

Migrating a profile lookup

// Proxycurl
const res = await fetch(
  "https://nubela.co/proxycurl/api/v2/linkedin?url=" + profileUrl,
  { headers: { Authorization: `Bearer ${key}` } },
);
const p = await res.json();

p.full_name
p.occupation
p.follower_count
p.city

// ScrapeField
const res = await fetch(
  `https://api.scrapefield.com/v1/linkedin/profile?url=${profileUrl}`,
  { headers: { Authorization: `Bearer ${key}` } },
);
const { data, meta } = await res.json();

data.full_name
data.headline
data.follower_count
data.location

The shape is close enough that most integrations are a rename. Three differences worth knowing before you start:

Everything is under data, with meta beside it carrying the request id, the credits charged and whether the answer came from cache.
Fields use LinkedIn’s own words: headline rather than occupation, and location as the profile prints it (“Berlin, Germany”) rather than split into parts.
A field we cannot see is null, never 0. And LinkedIn stops counting connections at 500, so connection_count is 500 for “500+” — we do not invent the rest.

Questions

Is LinkedIn data legal to buy at all?

Accessing data that is public and visible without logging in is the ground the whole industry stands on. The Proxycurl complaint was not simply “you scraped public pages” — it alleged fake accounts and non-public data, which is a different thing.

We access public, logged-out data only, and we never take your LinkedIn credentials. You are responsible for having a lawful basis for the personal data you request; see our acceptable use policy. None of this is legal advice.

How do I know you will not disappear too?

You do not, and anyone who tells you otherwise is selling. What we can tell you is the structure: LinkedIn is one of four surfaces and is capped at a minority of revenue, every surface is served by more than one source, and we hold no stored corpus of people to be sued over. The honest version is that a single-surface LinkedIn business is one lawsuit from zero, which is exactly why we are not one.

What about the fields Proxycurl had that you do not?

Experience, education and skills are on the profile, as the public page shows them. Anything we cannot get from a public, logged-out page we do not have and will not pretend to — if that breaks your use case, better to find out on this page than after you have integrated.

Try it before you decide

Every endpoint has a playground that runs against the live API with no key and no account — real routing, real validation, real errors. If the response shape does not fit your code, you will know in about a minute.

Open the docs See the credit table