Best LinkedIn Data APIs in 2026: The Real Cost Per 1,000 Records
If you need a LinkedIn data API in 2026 — company firmographics, job postings, or public profile basics — the honest answer to "which is cheapest" is not on any vendor's pricing page. Most sell you credits, and a credit is not a record. So we did the arithmetic ourselves.
On 13 August 2026 we opened every provider's live pricing page and converted each plan into the number that actually matters: effective cost per 1,000 LinkedIn records. The spread is much wider than the headline rates suggest — one entry plan advertised at "$40 for 200,000 credits" works out to $10.00 per 1,000 profiles, and one aggregated provider's own plan ladder spans a 39× range for the identical record.
We build one of these APIs, so we will be explicit about where our own LinkedIn API is a good fit and where it isn't. Every competitor figure below is dated and linked to the page we read it from.
TL;DR: There is no single "best" — it splits by workload. Cheapest entry point with no monthly commitment: Serpent API at $0.50 per 1,000 profile calls, then Bright Data at $1.50 pay-as-you-go. Bulk enrichment against a large historical database: Coresignal or Bright Data datasets, but only at the top of their plan ladders — their entry plans are the most expensive per record in this whole comparison. Flexible actor-based scraping: Apify at $4 per 1,000 profiles. And whatever you choose, wrap it in a thin adapter, because the last market leader here shut down with about eight weeks' notice.
Why the LinkedIn API market changed: the Proxycurl shutdown
For years a handful of vendors made pulling LinkedIn data feel like a solved problem, and the one most teams quietly hard-coded was Proxycurl, operated by Singapore-based Nubela. It is gone. This is the part of the story most "best LinkedIn API" round-ups still get vague about, so here is the documented record.
- 24 January 2025 — LinkedIn Corporation filed suit against Nubela Pte. Ltd. and Proxycurl LLC in the U.S. District Court for the Northern District of California, docket 3:25-cv-00828, before Judge Charles R. Breyer. The complaint bundles breach of contract, fraud, Computer Fraud and Abuse Act claims, unfair business practices and Lanham Act trademark claims. The docket is public on CourtListener.
- 4 July 2025 — Nubela published Goodbye Proxycurl: "In January earlier this year (2025), LinkedIn filed a lawsuit against Proxycurl. Today, we are shutting Proxycurl down." The stated reason was economic, not a court loss — under the American Rule a defendant generally cannot recover its legal fees even after winning, and the post notes that LinkedIn, owned by Microsoft, "has more or less an unlimited war chest."
- Scale of the loss — the same post describes Proxycurl as having reached "a ~$10M revenue business before we had to shut it down." This was not a hobby project quietly sunsetting; it was the category leader.
- Today (checked 13 August 2026) —
nubela.co/proxycurlnow serves a shutdown notice, "Proxycurl is no longer in service," and redirects to NinjaPear, a customer-data platform the founders moved on to. There is no API to sign up for and no migration path back.
Two things follow from this, and they should drive your choice more than any single feature.
Continuity risk is now a first-class buying criterion. A provider's answer to "will you be here next quarter" is worth more than a 20% discount. Ask where they operate, what data they touch, and whether their model depends on access that a platform can litigate away.
Switching cost is the only thing you actually control. Teams that had wrapped Proxycurl in a thin internal adapter swapped vendors in an afternoon. Teams that had scattered raw client calls across a codebase spent weeks. That is a design decision you can make today, for free, regardless of who you pick.
If the legal picture is what is really holding up your decision, we wrote a separate, carefully-hedged guide on whether scraping LinkedIn is legal in 2026, and a comparison of LinkedIn's official partner API versus scraping. If you are mid-migration right now, the Proxycurl alternative guide maps the old endpoints to their nearest equivalents.
What to look for in a LinkedIn data API
Before the tables, the criteria that actually matter:
- Coverage that matches your need. "LinkedIn data" is three different products: company firmographics, job postings, and person profiles. Very few tools are equally strong at all three. Decide which you truly depend on before optimising on price.
- Live scrape vs. aggregated dataset. Scraper APIs fetch a page on demand — fresh, but slower and page-dependent. Dataset providers match your query against a large pre-collected store — fast and bulk-friendly, but freshness varies per record. These are different products, not just different prices.
- Effective rate, not headline rate. This is the big one, and it gets its own section below. Per-call, per-record and per-credit models produce wildly different bills at your volume.
- Whether the plan floor exceeds your volume. A vendor whose cheapest plan is $1,500 a month is infinitely expensive if you need 5,000 records.
- A real free tier. You should be able to evaluate coverage on your own inputs before paying. Every provider here offers one.
- No login or OAuth required. Never hand a data vendor your LinkedIn credentials. Reputable providers work from publicly visible data only.
- Switching cost. After Proxycurl, assume any vendor could change. A thin adapter layer in your own code is the cheapest insurance available.
The comparison at a glance
Every price in this table was read off the vendor's own live pricing page on 13 August 2026 and is linked to its source. Vendors change these pages often — re-check before you commit.
| Provider | Best for | LinkedIn coverage | List pricing (read 13 Aug 2026) | Free to test | Model |
|---|---|---|---|---|---|
| Serpent API | Simple per-call company & jobs data | Company, jobs, public profile basics | $3.00/1K company; $0.50/1K profile, job search, job detail; 10% off at Growth, 30% off at Scale | 10 free calls, no card | Sync REST, one key |
| Bright Data | Enterprise-scale scraping & datasets | Profile, company, jobs, posts | Scraper API $1.50/1K pay-as-you-go; Scale $499/mo incl. 384,000 records, then $1.30/1K; LinkedIn dataset from $250 per 100K (~$2.50/1K) | 5,000 records/mo, no card | Sync + async batch, datasets |
| Apify (harvestapi) | Flexible, actor-based scraping | Profile, company, jobs, posts | $4/1K profiles; $10/1K profiles with email lookup; plans $0 / $29 / $199 / $999 per month | $5/mo credit, no card | Sync run + async, webhooks |
| Coresignal | Large aggregated professional dataset | Profiles, companies, jobs (aggregated) | Plans $49–$5,000/mo; employee or company record 10–20 credits, job posting 1 credit; bulk datasets from $1,000 | 2,000 credits, 7-day trial | API + bulk datasets + feeds |
| ScrapingDog | One credit bundle across many sites | Person profile, company page | Plans $40–$350/mo; a profile request costs 50 credits, or 100 if the profile is protected | 200 free credits, no card | Sync REST |
A note on freshness: Coresignal is an aggregated dataset provider — it matches your query against a large pre-built store, which is excellent for bulk enrichment but means an individual record can be older than a live scrape. Bright Data offers both a live scraper API and datasets. The rest are primarily live, per-request scrapers.
Other names you will meet while shopping include People Data Labs, ScrapIn, Piloterr and Nimble. They are legitimate, but their free tiers or published coverage limits made them harder to evaluate on equal terms, so we left them out rather than guess.
The real cost per 1,000: credits converted to records
Here is the thing nobody publishes. Three of the five providers above sell credits, and a LinkedIn record costs more than one credit. Until you divide, "200,000 credits for $40" tells you nothing about what a profile costs.
Method (so you can redo it). On 13 August 2026 we read each vendor's live pricing page for three numbers: the plan's monthly price, its included credit allowance, and the vendor's own published credit cost per record. Then:
effective $ per 1,000 = (plan price ÷ (credits ÷ credits per record)) × 1000
This assumes you consume the full monthly allowance. Credits generally do not roll over, so under-using a plan makes your real rate higher than the figure shown, never lower. Nothing here is estimated or interpolated — it is arithmetic on numbers each vendor publishes, and you can reproduce it in a spreadsheet in five minutes.
ScrapingDog: a profile costs 50 credits
ScrapingDog publishes credit bundles and states that "each Profile Scraper request costs 50 credits (100 if the profile is protected)." Dividing through:
| Plan | Price / month | Credits | Profiles at 50 credits | Effective $ / 1,000 profiles |
|---|---|---|---|---|
| Lite | $40 | 200,000 | 4,000 | $10.00 |
| Standard | $90 | 1,000,000 | 20,000 | $4.50 |
| Pro | $200 | 3,000,000 | 60,000 | $3.33 |
| Premium | $350 | 6,000,000 | 120,000 | $2.92 |
The entry plan is 3.4× more expensive per profile than the top plan, and $10.00 per 1,000 is well above every pay-as-you-go rate in this comparison. Protected profiles cost 100 credits, which doubles every number in that last column. This is not a criticism of ScrapingDog — the pricing is published honestly and the conversion is straightforward. It is a criticism of shopping by headline price.
Coresignal: a person or company record costs 10–20 credits
Coresignal publishes eight plan tiers and states that an employee or company record costs 10–20 credits while a job posting costs 1. Because the credit allowance grows far faster than the price, the effective rate collapses as you climb:
| Plan | Price / month | Credits | Records at 10 credits | Effective $ / 1,000 records (10 cr → 20 cr) |
|---|---|---|---|---|
| Mini | $49 | 2,500 | 250 | $196.00 → $392.00 |
| Starter | $199 | 12,000 | 1,200 | $165.83 → $331.67 |
| Pro | $499 | 35,000 | 3,500 | $142.57 → $285.14 |
| Growth | $1,000 | 150,000 | 15,000 | $66.67 → $133.33 |
| Premium | $1,500 | 1,000,000 | 100,000 | $15.00 → $30.00 |
| Scale | $3,000 | 4,000,000 | 400,000 | $7.50 → $15.00 |
| Elite | $5,000 | 10,000,000 | 1,000,000 | $5.00 → $10.00 |
That is a 39× spread for the same record ($196.00 down to $5.00 per 1,000, at 10 credits each) across one vendor's own menu. Two consequences are worth stating plainly:
- The entry plans cannot serve a 1,000-record job at all. Mini's 2,500 credits buy 125–250 employee records for the month. Starter's 12,000 buy 600–1,200. If your workload is a few thousand records, an aggregated provider is the wrong shape of product, not merely a pricier one.
- At the top, it is genuinely excellent value. Elite's $5.00–$10.00 per 1,000 records is competitive with anything, and job postings at 1 credit come out at $0.50 per 1,000 on that plan. Coresignal earns its reputation at volume; it just doesn't start there.
All five, on the same axis
Cheapest available rate versus what it costs to get in the door, for a single public profile record:
| Provider | Cheapest entry point (no monthly commitment) | Best rate available | What the best rate requires |
|---|---|---|---|
| Serpent API | $0.50 / 1K | $0.35 / 1K | One-time $500 deposit, no subscription |
| Bright Data | $1.50 / 1K (pay-as-you-go) | $1.30 / 1K | $499/mo Scale plan |
| Apify (harvestapi) | $4.00 / 1K (pay per event) | $4.00 / 1K | Flat — actor price does not tier |
| Coresignal | $196.00 / 1K ($49 Mini plan) | $5.00 / 1K | $5,000/mo Elite plan |
| ScrapingDog | $10.00 / 1K ($40 Lite plan) | $2.92 / 1K | $350/mo Premium plan |
The caveat that makes this table honest, and you should read it before quoting the numbers: these are not the same product. Cost per 1,000 is not cost per equivalent record. Serpent's profile endpoint returns publicly visible basics; Apify's $4 actor returns a detailed profile; Coresignal returns an enriched aggregated record with fields a live page scrape simply does not expose. A cheaper row can absolutely be the wrong buy. Price the fields you need — test each provider on your own inputs and count what comes back, which is exactly what we did when we documented which fields a LinkedIn profile API actually returns.
Provider by provider
A short, neutral read on each, grouped by how they work.
Live, per-request scraper APIs
Bright Data is the enterprise heavyweight: broad coverage across profiles, companies, jobs and posts, a scraper API at $1.50 per 1,000 pay-as-you-go, a $499/month Scale plan that includes 384,000 records and then bills $1.30 per 1,000, and a strong compliance posture. Its LinkedIn dataset is separately priced from $250 per 100,000 records — about $2.50 per 1,000 — against a store it advertises at 905 million-plus records. The free tier is the most generous here at 5,000 records a month with no card. It is powerful, and correspondingly more complex to adopt.
Apify is a marketplace of community-built "actors." Its widely-used harvestapi LinkedIn profile scraper uses pay-per-event pricing at a flat $4 per 1,000 detailed profiles, rising to $10 per 1,000 when you add email lookup — you are not billed for platform compute on top. Platform plans run $0 / $29 / $199 / $999 per month with $5 of monthly credit on the free tier. The trade-off is variance: reliability and field depth differ by actor, so benchmark the specific one you intend to run.
ScrapingDog sells one credit bundle usable across many targets, which is convenient if LinkedIn is one of several sites you scrape. Plans run $40 to $350 a month with 200 free credits to start. As shown above, the 50-credit-per-profile cost means the effective LinkedIn rate lands between $2.92 and $10.00 per 1,000 depending on plan.
Aggregated dataset providers
Coresignal serves a very large aggregated store of professional, company and job records through APIs, bulk datasets and feeds, with eight plan tiers from $49 to $5,000 a month and bulk datasets starting at $1,000. It is framed as public-web data rather than a branded LinkedIn scraper, which is the right mental model: excellent for breadth and bulk, less about live single-record freshness, and — per the table above — priced for buyers who are already at scale.
How to choose by use case
- Enrich a CRM of companies with firmographics: if you want transparent per-call pricing and don't need a huge historical lake, a per-request API like Serpent's company endpoint is a strong, cheap fit; there's a worked Python walkthrough in our guide to enriching leads with LinkedIn firmographics. For very large bulk backfills, a dataset provider is more economical per record — but only above the Premium tier, per the numbers above.
- Build a hiring-intelligence or jobs feed: you want a live jobs endpoint with pagination and real filters. Serpent, Bright Data and Apify all handle LinkedIn jobs well; compare at your daily volume, and note Coresignal's 1-credit job postings become very cheap at the top tiers. For the analysis layer on top, see tracking hiring trends from jobs data.
- Deep per-person profiles (full résumés): the hardest and most sensitive category. Dedicated profile products or an aggregated enrichment provider go deepest. Per-field completeness varies everywhere — test on your own inputs.
- Lowest possible switching cost: whichever you pick, put a thin adapter between your app and the provider so the vendor is a config value. That is the real lesson of the Proxycurl shutdown, and the same resilience mindset we apply to SERP scraping pipelines that face constant selector churn.
Where Serpent API fits (honestly)
We will be straight about our own tool. Serpent's LinkedIn API is the simple, per-call option: one REST key, no subscription, no minimum, and flat pricing — $3.00 per 1,000 company lookups and $0.50 per 1,000 profile, job-search or job-detail calls on the Default tier. A one-time $100 deposit unlocks Growth (10% off, so $2.70 and $0.45) and a one-time $500 deposit unlocks Scale (30% off, so $2.10 and $0.35). Those are deposits against usage, not recurring fees.
One clarification worth making, because it trips people up: our LinkedIn endpoints carry a 10% / 30% tier discount, while core web search carries a much steeper one — $0.60 per 1,000 on Default down to $0.06 at Growth and $0.03 at Scale. The same key covers our SERP and AI ranking APIs, and new accounts get 10 free calls to evaluate before paying anything.
Where we're strongest: company firmographics and job data at low-to-mid volume, where per-record vendors and monthly minimums hurt most. If you need a few thousand company records a month, the $49-to-$499 plan floor at an aggregated provider is the wrong shape and our $3.00 per 1,000 is hard to beat.
Where we're not the right pick: if you need deep, guaranteed per-person profile fields — skills, recommendations, certifications — or a billion-record historical lake to backfill against, an aggregated dataset provider or a dedicated profile product will serve you better. Our profile endpoint returns publicly visible basics and its depth varies by profile. We would rather tell you that here than have you discover it in production. Full numbers and response shapes are on the pricing page and in the API docs.
On continuity, the criterion this whole article opened with: we won't claim immunity from what happened to Proxycurl, because nobody honestly can. What we can tell you is that LinkedIn is one vertical of a broader search-data business, not the entire company, and that our pricing carries no lock-in for you to be trapped by — no subscription, no minimum, no expiring commitment.
Try a simple, per-call LinkedIn API
Serpent's LinkedIn API gives you company firmographics, public job search and detail, and best-effort public profile basics through one REST endpoint — from $0.50 per 1,000 calls, no subscription, with 10 free calls to evaluate. Strongest for company and jobs data; honest about profile scope.
Get Your Free API KeyExplore: LinkedIn API · Pricing · Documentation
FAQ
What is the best LinkedIn data API in 2026?
There is no single winner — it depends on what you pull and at what volume. For live, per-request scraping at low-to-mid volume, Serpent API and Bright Data are the cheapest entry points, at $0.50 and $1.50 per 1,000 records respectively as of 13 August 2026. For bulk enrichment against a large historical database, Coresignal and Bright Data datasets fit better, but only above their mid plan tiers. Apify is the pick when you want flexible, actor-based scraping you can customise.
What is the cheapest LinkedIn API?
Measured as effective cost per 1,000 records with no monthly commitment, Serpent API is cheapest at $0.50 per 1,000 profile, job-search or job-detail calls on the Default tier, followed by Bright Data at $1.50 per 1,000 pay-as-you-go and Apify at $4.00 per 1,000. Credit-based vendors look cheaper on the headline and are not: ScrapingDog's $40 Lite plan works out to $10.00 per 1,000 profiles at 50 credits each, and Coresignal's $49 Mini plan to $196.00 per 1,000 employee records at 10 credits each. All figures were computed on 13 August 2026 from each vendor's live pricing page.
Why does a credit-based LinkedIn API cost more than it looks?
Because a credit is not a record. ScrapingDog charges 50 credits per profile request and 100 for a protected profile; Coresignal charges 10 to 20 credits per employee or company record and 1 per job posting. To get a comparable number, divide the plan's credit allowance by the credits per record, then divide the plan price by that. Credits also generally do not roll over, so if you do not consume the full monthly allowance your real cost per record is higher than the published bundle implies.
What happened to Proxycurl?
LinkedIn Corporation sued Nubela Pte. Ltd. and Proxycurl LLC on 24 January 2025 in the U.S. District Court for the Northern District of California, docket 3:25-cv-00828. On 4 July 2025 Nubela announced it was shutting Proxycurl down, saying that even a win would not recover its legal fees. Checked on 13 August 2026, the Proxycurl site serves a shutdown notice and redirects to NinjaPear, an unrelated customer-data product. The API is gone and is not coming back — our Proxycurl alternative guide maps the old endpoints to replacements.
Do these APIs need a LinkedIn login or OAuth?
The reputable options here work with publicly available data and do not ask for your LinkedIn credentials or an OAuth grant. Serpent API's LinkedIn endpoints require neither — you call a REST endpoint with your Serpent key. Aggregated providers like Coresignal serve from their own datasets rather than a live login at all. If a vendor asks you to supply LinkedIn session credentials, treat that as a red flag.
Is it legal to use a LinkedIn data API?
The law around scraping public web data is unsettled and fact-specific, and this article is not legal advice. LinkedIn's complaint against Nubela raised contract, fraud, CFAA and trademark theories rather than one clean question, which is precisely why the case never produced a tidy answer. Working only with public data and a reputable provider lowers risk, but you remain responsible for using any data lawfully and within the terms that apply to you. Consult counsel for your specific use case.
