Quickstart · Command · Endpoints · Data · Errors · Coding agents · Docs · Support
LinkedIn profiles, companies, jobs and posts as JSON, in Python. No LinkedIn login, no browser. Built on the Bright Data LinkedIn Scraper API.
Uses the Bright Data Python SDK. Full API docs: LinkedIn Scraper API.
Also here: a one-command CLI for profiles, and the Bright Data CLI, which needs no Python at all.
LinkedIn profile data is personal data. Use it within the law that applies to you: Bright Data compliance.
Python 3.10 or newer.
pip install brightdata-sdk
export BRIGHTDATA_API_TOKEN=YOUR_API_KEYGet a token from the
Bright Data control panel. A .env
file in the project root works too, as long as the virtualenv is inside the
project.
Or skip the token. Run npx -p @brightdata/cli bdata login once: it opens a
browser, and from then on the SDK finds the stored credentials on its own, for
you and for any coding agent working in that terminal. Agents cannot click
through the login, so do it yourself first.
No account yet? Create one; new accounts get 5,000 free credits a month.
from brightdata import SyncBrightDataClient
with SyncBrightDataClient(auto_create_zones=False) as client:
profile = client.scrape.linkedin.profiles("https://www.linkedin.com/in/satyanadella/").data
print(profile["name"], "|", profile["position"])
print(profile["followers"], "followers,", profile["connections"], "connections")Satya Nadella | Chairman and CEO at Microsoft
12173313 followers, 500 connections
Expect under a minute: the API runs a job and the SDK waits for it. One credit per profile.
Pass auto_create_zones=False every time. Left on, the SDK creates zones on
startup for Web Unlocker and SERP, two other Bright Data products this scraper
never touches, and zone creation fails on accounts without a payment method
(sdk-python#57).
The command in this repo does the same for several profiles and writes one JSON file.
pip install git+https://github.com/brightdata/linkedin-scraper-python
linkedin-scraper satyanadella reidhoffmanFetching 2 LinkedIn profiles: satyanadella, reidhoffman
One job for all of them, usually one to three minutes. One credit per profile.
asking 2 profiles...
got satyanadella: 34 fields (Satya Nadella)
got reidhoffman: 38 fields (Reid Hoffman)
Saved 2 of 2 profiles as JSON to linkedin.json
It takes a profile URL just as happily as the slug inside it.
In a terminal the asking line is replaced by this, updating in place, so you
can see it is working and how long it has been going:
⠹ 2 profiles 0:00:41
--out PATH output file, default linkedin.json
python -m linkedin_scraper works too.
Every profile you ask for goes into one job. The API bills per record, not per job, and a job takes about the same time for one URL as for ten.
Import it instead of running it, for ok and error per profile instead of raw
rows. scrape never raises for one bad profile; check ok before reading
profile:
from linkedin_scraper import scrape
for outcome in scrape(["satyanadella", "zz-not-a-real-profile-zz"]):
if outcome.ok:
print(f"{outcome.slug}: {outcome.profile['name']}")
else:
print(f"{outcome.slug} failed: {outcome.error}")satyanadella: Satya Nadella
zz-not-a-real-profile-zz failed: The profile is hidden or private.
The command covers the first row of the table below. The rest of the SDK's
LinkedIn surface is the other rows, documented in the
LinkedIn Scraper API docs.
Every snippet below is complete and needs only brightdata-sdk: paste it as
is. Every one of them runs in Actions each Monday, a smaller check runs every
other day, and the badge at the top is the latest result.
| you have | want | call |
|---|---|---|
| a profile URL | that profile | client.scrape.linkedin.profiles(url) |
| a first and last name | matching profiles | client.search.linkedin.profiles(first_name, last_name), see the note below |
| a company URL | that company | client.scrape.linkedin.companies(url) |
| a job URL | that job posting | client.scrape.linkedin.jobs(url) |
| a keyword or a location | matching job postings | client.search.linkedin.jobs(keyword=..., location=...), see the note below |
| a post URL | that post | client.scrape.linkedin.posts(url) |
| a profile or company URL | its posts | client.search.linkedin.posts(url, start_date=..., end_date=...) |
search.linkedin.profiles takes a first and last name, not a URL. It is people
search by name. This README documents that it exists and does not demonstrate
it against a real person.
search.linkedin.jobs has no parameter that caps how many postings it returns,
and it bills one credit per posting. A broad keyword matches thousands. It is
not run here, because every snippet in this README reruns weekly. Narrow it with
company, location and timeRange before running it wide.
There is no example here built on a pinned job URL. A posting closes, and a snippet pinned to it would turn red the week that happened.
Every one of these is an asynchronous job. The SDK triggers it, polls, and returns when it is ready. The API's synchronous endpoint, 20 URLs and a one-minute limit, is raw HTTP only.
Error rows, like the dead profile above, appear because the SDK asks for them
with include_errors=true. The API default is off.
scrape calls with one URL return one record as a dict, and several as a list.
The timeout argument is seconds, 180 by default.
A list of URLs is one job, not one per profile.
from brightdata import SyncBrightDataClient
with SyncBrightDataClient(auto_create_zones=False) as client:
results = client.scrape.linkedin.profiles([
"https://www.linkedin.com/in/satyanadella/",
"https://www.linkedin.com/in/reidhoffman/",
])
for result in results:
row = result.data
source = row.get("input_url") or row["input"]["url"] # an error row has only input
print(source, "|", row.get("name") or row["error"])https://www.linkedin.com/in/reidhoffman/ | Reid Hoffman
https://www.linkedin.com/in/satyanadella/ | Satya Nadella
The URLs went in as satyanadella then reidhoffman, and the rows came back the
other way round. Each line is still right, because it reads the row's own
input_url.
Read which profile a row is from its own input_url, as above, never from
result.url. The SDK pairs results with the URLs you passed by position, and
the API returns rows in a different order each run, so result.url can name
one person while result.data holds another's profile. Every result is also
marked success=True, including a dead profile's error row
(sdk-python#60). An error
row has no input_url; it names its profile in input["url"].
For anything bigger than a few profiles, do not block a process for an hour. Trigger, keep the snapshot id, fetch when ready. Snapshots stay downloadable for 30 days.
import time
from brightdata import SyncBrightDataClient
with SyncBrightDataClient(auto_create_zones=False) as client:
job = client.scrape.linkedin.profiles_trigger("https://www.linkedin.com/in/satyanadella/")
print("snapshot:", job.snapshot_id)
while (status := client.scrape.linkedin.profiles_status(job.snapshot_id)) not in ("ready", "failed"):
time.sleep(5)
print("status:", status)
record = client.scrape.linkedin.profiles_fetch(job.snapshot_id)[0]
print("fetched:", record["name"], "|", record["current_company_name"])snapshot: sd_mu717o9x2md4afafsv
status: ready
fetched: Satya Nadella | Microsoft
from brightdata import SyncBrightDataClient
with SyncBrightDataClient(auto_create_zones=False) as client:
company = client.scrape.linkedin.companies("https://www.linkedin.com/company/bright-data/").data
print(company["name"], "|", company["employees_in_linkedin"], "employees on LinkedIn")Bright Data | 407 employees on LinkedIn
The fields most people want from a profile:
name position city current_company_name followers connections experience
The code hardcodes no field list. Whatever the API returns lands in
result.data, and in the command's file.
The API marks 8 of these fields as personal data, with pii: true in the
schema: id, name, about, url, input_url, linkedin_id, first_name
and last_name.
get_metadata() drops that flag. It keeps type, active, required and
description for each field and discards the rest
(sdk-python#61). To read
it, request https://api.brightdata.com/datasets/gd_l1viktl72bvl7bjuj0/metadata
yourself with the same token. The table below does that.
All 47 fields, with type and description
Regenerated every day from the dataset schema, via the raw metadata
endpoint, so it cannot go stale. A
profile carries the fields that apply to it: the sample file has 32
of these 47, plus timestamp and input,
which the schema does not list.
| field | type | description |
|---|---|---|
id |
text | Personal data. A unique identifier for the person's LinkedIn profile |
name |
text | Personal data. Profile name |
city |
text | Geographical location of the user |
country_code |
text | Geographical location of the user |
position |
text | The current job title or position of the profile |
about |
text | Personal data. A concise profile summary. In some cases, only a truncated version with "…" is displayed on the website, and this is the version we capture |
posts |
array | Contains information related to the user's last LinkedIn posts. It typically includes the post title, created date, URL link to the post, etc. |
groups |
array | The LinkedIn groups that the profile is a part of |
current_company |
object | Provides information about the user's current professional position. It typically includes the company name, the user's job title, the company ID, and the industry or sector to which the company belongs |
experience |
array | Contains information about user's professional history. It typically includes the user's job title, length of time the user held the position, the geographic location of the company, the start and end date, the company name, URL link to the company profile, etc. |
url |
url | Personal data. URL that link directly to the LinkedIn profile |
people_also_viewed |
array | Provides a list of LinkedIn profiles that users who have viewed the user's profile, have viewed these as well |
educations_details |
text | Provides information about the user's educational background |
education |
array | Provides information about the user's educational background. It typically includes the degree, the start and end year, the filed, etc. |
recommendations_count |
number | A numeric count of the total number of recommendations that the user has received |
avatar |
url | URL that link to the profile picture of the LinkedIn user |
courses |
array | Contains information about courses or educational programs that the user has undertaken |
languages |
array | Contains information about the user's proficiency in different languages |
certifications |
array | Licenses & Certifications |
recommendations |
array | Recommendations that the user has received from their connections or colleagues on LinkedIn |
volunteer_experience |
array | Contains information related to the user's volunteer work |
followers |
number | How many users/ companies following the profile |
connections |
number | How many connections the profile has |
current_company_company_id |
text | The id of the latest/current company of the profile |
current_company_name |
text | The name of the latest/current company of the profile |
publications |
array | Published works or presentations |
patents |
array | Patents filed or granted |
projects |
array | Professional or academic projects |
organizations |
array | Memberships in professional organizations |
location |
text | Geographical location of the user |
input_url |
url | Personal data. The URL that was entered when starting the scraping process |
linkedin_id |
text | Personal data. LinkedIn profile identifier |
activity |
array | Any activity the user has regarding posts |
linkedin_num_id |
text | Numeric LinkedIn profile ID |
banner_image |
url | Banner image |
honors_and_awards |
array | Awards and recognitions received |
similar_profiles |
array | Profiles similar to the current one |
default_avatar |
boolean | Is the avatar picture the default avatar empty picture |
memorialized_account |
boolean | Boolean indicating if the account is memorialized |
bio_links |
array | External links added to the bio |
first_name |
text | Personal data. First name of the user |
last_name |
text | Personal data. Last name of the user |
urn_id |
text | The Uniform Resource Name (URN) used by LinkedIn |
urn |
text | Uniform Resource Name |
influencer |
boolean | Indicator if the profile marked as influencer |
fsd_profile_id |
text | FSD profile ID |
backfilled_columns |
object | Indicates whether sticky columns were backfilled. Keys are column names and values are true/false. |
The start of a real output file, from linkedin-scraper satyanadella
{
"generated_at": "2026-09-18T14:15:08.047993+00:00",
"profiles": [
{
"slug": "satyanadella",
"profile": {
"id": "satyanadella",
"name": "Satya Nadella",
"city": "Redmond, Washington, United States",
"country_code": "US",
"position": "Chairman and CEO at Microsoft",
"about": "As chairman and CEO of Microsoft, I define my mission and that of my company as empowering every person and every organization on the planet to achieve more.",
"posts": [
{
"title": "How do we build a frontier intelligence ecosystem?",
"attribution": "Great to be back at Microsoft Build today. For us, it is not about any one piece of technology or even the platform.",
"img": "https://media.licdn.com/dms/image/v2/D560DAQHI1Iu8CrO8CQ/learning-public-crop_288_512/B56ZvrnmmhJ8AQ-/0/1769184586791?e=2147483647&v=beta&t=IZT3h7JK8tLoX5KYgwA0xLjxIKtc14hmHqKCk1cOLP8",
"link": "https://www.linkedin.com/pulse/how-do-we-build-frontier-intelligence-ecosystem-satya-nadella-73jhc",
"created_at": "2026-06-02T00:00:00.000Z",
...The whole file, one profile with every field, is examples/sample_output.json.
| you see | what it means |
|---|---|
API token required but not found. |
Exit 2, before any request. Set the token. |
failed slug: The profile is hidden or private. |
Exit 1. No such profile, or not public. Usually a typo in the slug. |
failed slug: the API returned no row for this profile |
Exit 1. The job came back without a row for that input. Run it again. |
failed slug: timeout |
Exit 1. A request gives up after 180 seconds. Run it again. |
failed slug: Failed to fetch results: Request timeout after 30 seconds |
Exit 1. The job finished, but downloading it took longer than 30 seconds. Run it again. |
failed slug: Crawler error: Your system is sending too many of this type of request. ... |
Exit 1. The account is being throttled, and a valid profile fails with it. Wait, then run it again. It clears on its own. |
Any failure exits 1, so a run is safe to gate a script on.
From the SDK, the same conditions look like this:
| you see | what it means |
|---|---|
AuthenticationError: Unauthorized (401) |
The token is set but wrong. |
result.success is False, result.status is "timeout" |
The SDK gave up waiting for the job, 180 seconds by default. Pass timeout=420 to the call, or run it again. |
Failed to fetch results: Request timeout after 30 seconds |
The job finished; downloading it did not. That is the client's own limit on each HTTP request, SyncBrightDataClient(timeout=30), not the timeout you pass to the call. Raising the call's timeout does not help. Run it again, or pass a larger timeout to the client. |
a dict in result.data with an error key |
The API's answer for one input. result.success still says True; read the row. |
No Python, nothing installed. Paste both lines; the first opens a browser
once, or use bdata login --device over SSH and in CI:
npx -p @brightdata/cli bdata login
npx -p @brightdata/cli bdata pipelines linkedin_person_profile "https://www.linkedin.com/in/satyanadella/"bdata pipelines list prints every type. The LinkedIn ones are
linkedin_person_profile, linkedin_company_profile, linkedin_job_listings,
linkedin_posts and linkedin_people_search. Each takes URLs, prints JSON,
and costs one credit per record.
npx skills add brightdata/skills teaches Claude Code, Cursor and Codex these
commands and the docs, so plain language works afterwards. Full guide:
Bright Data for your coding agent.
No terminal, for a hosted assistant? The
Bright Data MCP server
has LinkedIn tools in its social group, which is off unless you ask for it:
https://mcp.brightdata.com/mcp?token=YOUR_API_TOKEN&groups=social
An agent can also open the account itself, no signup form: agent registration. Everything else Bright Data connects to, from LangChain to Zapier and n8n: integrations.
Bugs in this repo: open an issue, and CONTRIBUTING.md says what to put in it. Anything about the API, your account or your credits: Bright Data support.
MIT.