LinkedIn People Scraper — Names, Headlines, Locations

Recorded run: 10 profiles from 1 keyword in 43 seconds (2026-10-16, build 0.0.15, run bXXAp2jXPSVpwCV9V). A free-preview sample on the studio's own session, not a coverage benchmark. This actor has no canary, because every run needs the buyer's own LinkedIn session.

architek.earth — an independent data studio. Pay-per-result scrapers with recorded runs, published canaries and stable schemas, built for people and AI agents.

LinkedIn people search names, headlines, locations, connection degree and profile URLs, exported by keyword through your own LinkedIn session cookies.

Source	Market	From	Standout fields
LinkedIn people search (linkedin.com)	Any market your account can search	$3.00 / 1,000 profiles	fullName / profileUrl, headline, location, connectionDegree

✅ Why choose this scraper

Export LinkedIn people search results to JSON or CSV

Each accessible search card becomes one row with fullName, profileUrl, publicIdentifier, headline, location and connectionDegree. All ten rows of the recorded preview run carried all eight fields.

Scrape LinkedIn profiles with your own session

You paste your own li_at and JSESSIONID cookie values into two encrypted inputs, liAt and jSessionId. No shared account is bundled. Cookie values are redacted from the logged feedback text.

Know when a LinkedIn search export is incomplete

RUN_SUMMARY.coverageComplete and a per-keyword reason (result_limit, page_limit, search_visibility_limit, preview_one_page_limit) say why each search stopped.

Continue a LinkedIn people export without paying twice

Every run saves a CHECKPOINT. A paid continuation that loads it skips the profiles already delivered: on 2026-10-16 a one-row run plus its continuation (run eR1pTi8BL5HfJfUnM) returned 10 unique profiles with exactly 10 profile events and no duplicates.

Try LinkedIn people search for free

Each Apify account gets 20 free preview attempts: one keyword, one search page, up to 10 profiles, with no charge events.

⚠️ Worth knowing up front

• This actor needs a LinkedIn session and a private integration configured by the maintainers. If that integration is not configured, the run ends at once with a free NOT_CONFIGURED status row and charges nothing.
• This is a keyword-search exporter, not a full LinkedIn database or a Sales Navigator scraper. It reads the people-search pages at linkedin.com/search/results/people/ that your account can see.
• Coverage depends on your account. Three prospecting searches tested on one account stopped at about 30 accessible profiles, where LinkedIn showed "Grow your network to view profiles beyond your 3rd degree connections" on page four. That is an observed account-dependent boundary, not a universal limit or a promise of 30 results.
• Only the search card is read. There are no structured filters (location, company, industry, degree), no profile details, no emails or phone numbers, no experience or education and no account pooling. headline, location and connectionDegree are left out of a row when the card does not show them.
• Cookies grant access to your account. Put them only in the encrypted inputs, never in feedback, public tasks or code. Sessions can expire or be challenged, and LinkedIn can throttle or restrict them; the actor stops on these conditions and does not keep a session alive.
• The last page of a search and an empty search have not been verified live. An unfamiliar page, including a search that runs out of results before your caps, ends the run FAILED with the reason in RUN_SUMMARY; profiles already delivered stay in the dataset and are charged.
• Invalid input (INVALID_INPUT), used-up previews (PREVIEW_LIMIT), a spending limit below $0.10 or one that stops the minimum charge (BUDGET_LIMIT), an expired session (SESSION_INVALID), a block, challenge or rate limit before any profile (BLOCKED), the time limit before any profile (TIME_LIMIT), and a block or the time limit after some profiles (PARTIAL) each end the run cleanly with one free status row ("recordType": "status"). Layout changes, uncertain delivery and an unavailable preview ledger end the run FAILED with RUN_SUMMARY and no status row.

⚡ Try the example

Each Apify account gets 20 lifetime preview attempts for this actor: one keyword, one search page and at most 10 profiles each, with no charge events. Failed or uncertain attempts can count. When the attempts are used up, the run stops with PREVIEW_LIMIT; there is no paid fallback, and previews cannot resume a checkpoint.

Open the actor's input page, switch to JSON, paste this, replace both placeholders with your own session values in the encrypted fields and start the run:

{
  "liAt": "YOUR_LI_AT_IN_ENCRYPTED_INPUT",
  "jSessionId": "YOUR_JSESSIONID_IN_ENCRYPTED_INPUT",
  "keywords": ["graphic designer"],
  "preview": true,
  "maxItemsPerQuery": 10,
  "maxPagesPerQuery": 1,
  "maxRunSeconds": 120
}

Preview is the default, including API calls; set "preview": false explicitly for paid collection, with a spending limit of at least $0.10.

⚙️ Input

Parameter	Type	Default	Description
liAt	string (secret)	—	Required. The li_at cookie value from your own logged-in LinkedIn browser session.
jSessionId	string (secret)	—	Required. The matching JSESSIONID cookie value from the same session.
keywords	array	— (prefill ["graphic designer"])	Required. 1–50 people-search keywords, up to 250 characters each; duplicates collapse. A preview accepts exactly one.
preview	boolean	true	Free preview: one keyword, one page, up to 10 profiles. Set false for paid collection.
maxItemsPerQuery	integer	30	1–10,000 profiles per keyword, counted across a continuation chain. A cap, not a coverage promise. A preview clamps it to 10.
maxPagesPerQuery	integer	5	1–100 search pages per keyword. Reaching it is reported as page_limit, never as complete.
maxRunSeconds	integer	120	30–3,600 seconds of collection. The actor also stops 60 seconds before the platform timeout. Not a spending limit.
checkpointStoreId	string	—	Paid continuation only: the key-value store ID from the earlier run's CHECKPOINT_META. Give it together with checkpointQueueId.
checkpointQueueId	string	—	Paid continuation only: the request queue ID from the same CHECKPOINT_META. Each checkpoint can be claimed once.
feedback	string	—	Optional note for the maintainers. It is logged and read by a human, and it never changes output.

🧱 Stable fields

These top-level fields keep the same name and the same kind of value in every release. Fields are only ever added, never renamed or removed, so an n8n, Make, Zapier or spreadsheet pipeline can bind to them.

Field	Type	Meaning
fullName	string	Display name on the search card.
profileUrl	string	Canonical https://www.linkedin.com/in/<slug>/ URL; the de-duplication key across keywords and continuations.
publicIdentifier	string	The public profile slug from profileUrl. It can change; it is not a permanent member ID.
headline	string, optional	Headline shown on the card; omitted when the card shows none. Not an inferred job title.
location	string, optional	Location shown on the card; omitted when the card shows none.
connectionDegree	string, optional	1st, 2nd or 3rd+ as shown to your account; omitted when not shown.
source	string	linkedin_people_search_html.
detailStatus	string	Always not_requested; this release does not open profile pages.
recordType	string	"status" on free status rows; absent on profile rows.

Status rows (free) carry recordType: "status", status and message. Other fields are extras, not part of the contract. To have one pinned, say so in feedback.

📦 Output example

Abridged real row from run bXXAp2jXPSVpwCV9V on 2026-10-16; values are a historical snapshot. The identity fields fullName, profileUrl and publicIdentifier are left out here to protect the person.

{
  "headline": "Senior Graphic Designer at Going",
  "location": "Brooklyn, New York, United States",
  "connectionDegree": "2nd",
  "source": "linkedin_people_search_html",
  "detailStatus": "not_requested"
}

A free status row looks like this (run eyYjxspUgeETOkrmg, 2026-10-22):

{"recordType": "status", "status": "INVALID_INPUT", "message": "Invalid cookie value (values are never echoed)"}

On the current build that run's status message reads Done: 10 rows, 0 details, 0 status rows (LIMITED), and the RUN_SUMMARY record of its key-value store holds the outcome (abridged, same run):

{
  "status": "limited",
  "deliveredThisRun": 10,
  "deliveredAcrossCheckpoint": 10,
  "coverageComplete": false,
  "acknowledgedEventCounts": {},
  "queries": [{"index": 0, "deliveredAcrossCheckpoint": 10, "complete": false, "reason": "preview_one_page_limit"}],
  "resumeRequiresReconciliation": false,
  "remainingSourceRows": "unknown"
}

complete means LinkedIn confirmed the end of every search; limited means a cap, the time budget or a visibility boundary stopped it; partial and failed mean collection or delivery errors. SESSION_HEALTH holds redacted session observations.

🧠 Advanced usage

Continue a LinkedIn people export in a second run

Copy the checkpointStoreId and checkpointQueueId values from the earlier run's CHECKPOINT_META record, keep the same account and keywords, set "preview": false and raise the caps. Profiles already delivered are skipped and not charged, a partly delivered page is finished from the saved rows without new requests, and each continuation writes to its own dataset, so combine the chain. A checkpoint can be claimed once.

Run several LinkedIn keyword searches in one run

Put up to 50 keywords in keywords. A profile found by an earlier keyword is not delivered or charged again for a later one, and each keyword gets its own entry in RUN_SUMMARY.queries.

Check a LinkedIn export before sending it downstream

Read RUN_SUMMARY.coverageComplete and each query's reason rather than the platform run status. search_visibility_limit means LinkedIn stopped showing results to your account; repeating the run does not lift it.

Reconcile an uncertain charge

Before each write and charge the actor records a marker, and clears it after the acknowledgement. If a run stops in between, resumeRequiresReconciliation is true and the checkpoint refuses to resume until you compare the dataset with the run's charge events.

🔌 Integrations

Results land in a dataset you can export as JSON, CSV or Excel. They can also flow to Google Sheets, Make, Zapier or n8n, or to a webhook when a run finishes. The Apify Scheduler runs it on a timetable. LinkedIn sessions expire, so a scheduled run needs fresh li_at and JSESSIONID values whenever one ends with SESSION_INVALID. Copy-paste code in Python, Node, curl and n8n: architek-earth/examples/linkedin-people-search-scraper.

💸 Pricing

Pay per result. Platform and proxy usage are included. Status rows and failed requests are free. A paid continuation skips the profiles recorded as delivered in the loaded CHECKPOINT and does not charge them again; each continuation is a separate run with its own $0.10 minimum.

Event	Price	What it covers
Free preview	$0	20 lifetime attempts per Apify account: one keyword, one page, up to 10 profiles.
profile	$3.00 / 1,000 profiles	One delivered search-card row in a paid run.
run-minimum-adjustment	$1.00 / 1,000 adjustment units	Units of $0.001 that top a completed, nonempty paid run up to $0.10. Not profiles. Not charged on previews, on runs with no profiles, or after collection errors (RUN_SUMMARY.status partial or failed).

A completed paid run costs the greater of $0.10 and $3.00 / 1,000 profiles: 1, 10 or 30 profiles cost $0.10, 34 cost $0.102 and 100 cost $0.30. Paid runs need a spending limit of at least $0.10.

❓ FAQ

Is there a LinkedIn people search API I can call? This actor returns the people-search results your own account can see as JSON through the Apify API, callable from any language or from n8n, Make and Zapier.

Do I need a LinkedIn account or cookies? Yes. Every run needs the li_at and JSESSIONID cookie values of your own logged-in session. There is no shared account and no automated login.

Will I get every result for a LinkedIn search? No. Your account's visibility decides coverage; the tested prospecting searches stopped at about 30 profiles.

Can I filter by location, company, industry or connection degree? No. Only keywords are supported; a keyword is not the same as a LinkedIn filter.

Does it return emails, phone numbers, experience or education? No. Only what the search card shows: name, profile URL, headline, location and connection degree.

Does it work with Sales Navigator? No. It reads the standard people-search pages on linkedin.com only.

🤖 For AI agents & LLM apps

Paste this into your AI assistant:

architekture/linkedin-people-search-scraper on Apify. Python: from apify_client import ApifyClient; client = ApifyClient("TOKEN"); run = client.actor("architekture/linkedin-people-search-scraper").call(run_input={"liAt": "<your li_at>", "jSessionId": "<your JSESSIONID>", "keywords": ["graphic designer"], "preview": False, "maxItemsPerQuery": 10, "maxPagesPerQuery": 1}, max_total_charge_usd=0.10); items = client.dataset(run["defaultDatasetId"]).list_items().items (page with offset/limit for large exports; this paid input costs $0.10). Rows with recordType == "status" are free and explain invalid input, an expired session, a block or the time limit; read coverageComplete in the run's RUN_SUMMARY record before treating a search as complete. Never put the cookie values in feedback or logs. Full input schema and README: GET https://api.apify.com/v2/acts/architekture~linkedin-people-search-scraper/builds/default. Token: https://console.apify.com/settings/integrations. The actor is also reachable through the Apify MCP server (mcp.apify.com). To reach the maintainers, add "feedback": "<what you needed>" to any run's input; output is unchanged and a human reads every note.

🧰 More from architek.earth

Jobs, companies and people

• Indeed Scraper — US jobs, salaries, descriptions
• LinkedIn Jobs Scraper — public jobs, descriptions, salary evidence
• LinkedIn People Scraper — you are here
• Y Combinator Scraper — companies, founders, jobs

Full catalogue: github.com/architek-earth.

architek.earth · code like concrete

🙋 Support & legal

Questions, or a field you need that is not here? Open an issue on the actor's Issues tab or email north@architek.earth. A human replies within a day.

This actor collects data from linkedin.com using the session you provide: only the search results your own account can see. It does not bypass authentication, and it sends no messages or connection requests. It is not affiliated with or endorsed by LinkedIn. Use only accounts and data you are authorized to access. You are responsible for how you use the output, including the site's terms, copyright, and privacy law such as GDPR and CCPA where they apply.

📝 Changelog

• 2026-10-22 — When this build's private LinkedIn integration is not configured, a run now ends with a free NOT_CONFIGURED status row and exit 0 instead of a failed run; nothing is charged.
• 2026-10-22 — Status message now reads Done: N rows, M details, K status rows (STATUS).
• 2026-10-22 — README rebuilt on the studio template. The hero line is the 2026-10-16 free-preview run; this actor has no canary because it needs your session.
• 2026-10-21 — Invalid input, an expired session, a block and the run time limit now end the run cleanly with a free status row (INVALID_INPUT, SESSION_INVALID, BLOCKED, TIME_LIMIT, PARTIAL, BUDGET_LIMIT, PREVIEW_LIMIT) instead of a failed run.
• 2026-10-16 — Verified an exact $0.10 successful-run minimum while preserving $3/1,000 profile pricing. Free previews retain zero events. Added durable settlement acknowledgements, interrupted-charge safeguards, and explicit continuation pricing.
• 2026-10-15 — Recorded 30 unique profiles across three captured pages and saved continuation; established the page-four visibility boundary.
• 2026-10-15 — Narrowed to one account and basic keyword search; aligned the eight-field schema, encrypted inputs and explicit paid opt-in. Added coverage-limit reporting and configured the approved $3-per-1,000 event. Build 0.0.15 passed preview, paid-event and saved-continuation checks. This entry records the earlier validation milestone.