LinkedIn Post Search

🔎 LinkedIn Post Search API: Search LinkedIn Posts by Keyword from Python and MCP

LinkedIn post search by keyword, name, phrase or #hashtag: get every matching public post as JSON, with the full text, author, date, reactions, comments, images and links. No LinkedIn login, no cookie.

• Actor: LinkedIn Post Search API on Apify
• Input schema: input parameters
• Get a free API token: apify.com

This repo is a Python and MCP quick-start for the LinkedIn Post Search API on Apify. Give it keywords, names or hashtags and a date window such as the past day or the past week, and it returns the public LinkedIn posts that mention them, one row per post. It is the search LinkedIn itself only offers to logged-in members, run as an API over the public side of LinkedIn: no account, no session cookie, nothing attributed to you. Run it on a schedule and it becomes a LinkedIn social listening feed; run it once with a date range and it is research for a story, a launch or an event.

Coverage is the public, indexed part of LinkedIn. Posts from members who keep their activity private, and posts published in the last few hours, may be missing, and a signed-out visitor sees at most 10 comments per post. For a mention monitor that is usually what you want: the posts that are public and spreading.

Responsible use: the API returns only what LinkedIn shows signed-out visitors and collects no contact details, and you are responsible for using the data lawfully.

Video Walkthrough

Text walkthrough

The LinkedIn Post Search API turns a keyword into a list of public LinkedIn posts. You pass keywords (up to 20 names, phrases or #hashtags, each matched as an exact phrase), pick postedWithin (day, week, month, year) or a startDate and endDate, and cap the run with maxPostsPerKeyword. Each row carries text, authorName, authorUrl, authorType, datePosted, numLikes, numComments, topComments, hashtags, images, embeddedLinks, taggedPeople and taggedCompanies, plus matchedKeyword and the search page and rank it came from. For LinkedIn hashtag analytics, pass #semiconductors with postedWithin: "week" and sort the rows by numLikes to see which posts and authors lead the tag; for brand monitoring, pass your company and your executives with postedWithin: "day" on a daily schedule. The default run in this repo asks for five posts about one event and costs well under a cent; posts and search pages read in the last 7 days come back from a shared cache, so repeats are cheaper and faster.

Quick Start

Prerequisites

• Python 3.11 or higher
• An Apify account and API key (get a free key here)

1. Clone the repository

   git clone https://github.com/johnisanerd/Apify-LinkedIn-Post-Search-API.git

cd Apify-LinkedIn-Post-Search-API

1. Install dependencies with UV

   # Install UV if you do not have it:

curl -LsSf https://astral.sh/uv/install.sh | sh

Install project dependencies:

uv sync

1. Configure your API key

   cp .env.example .env

Edit .env and add your Apify API key

Get your free API key at: https://apify.com?fpr=9n7kx3

1. Run the example

   uv run python linkedin-post-search-api-example.py

The script starts one run for "Jackson Hole Symposium" with maxPostsPerKeyword: 5, waits for it to finish, prints each post with its author, date, reactions and link, and then prints the run's RUN_SUMMARY (pages fetched, cache hits, why each keyword stopped). Three recipe helpers mirror the use cases on the Actor's Examples tab:

uv run python linkedin-post-search-api-example.py hashtag    # track #semiconductors this week

uv run python linkedin-post-search-api-example.py mentions # past-day brand and executive mentions uv run python linkedin-post-search-api-example.py top # most liked posts on a topic this week

Alternative: set the API key directly

export APIFY_API_TOKEN="your_api_key_here"

uv run python linkedin-post-search-api-example.py

Why Use This LinkedIn Post Search API?

No LinkedIn login or cookie

The API reads only what LinkedIn shows a signed-out visitor. There is no field for an account, a password or an li_at cookie, so no LinkedIn account is ever at risk, and anyone with a free Apify token can search LinkedIn posts without an account.

A date window that is checked, not guessed

postedWithin and the startDate / endDate range are applied to the search and then checked against every post's own timestamp before it is delivered, so you are never billed for a post outside the window. LinkedIn's own feed cannot filter posts by date; this API can.

Engagement and context on every post

Each row carries reactions, comment count, up to 10 of the comments a signed-out visitor sees, the people and companies tagged in the post, the hashtags, the media and links, and the original post when the match is a share.

A shared cache that makes monitors cheap

Posts and search pages read in the last 7 days are reused across runs. Cached search pages are free, and a daily monitor pays only for the pages and posts that are new.

Pay per search page and per post, no start fee

One small event per page of search results fetched fresh and one per post delivered. Error rows, posts outside the window and cached pages cost nothing.

Recipes

These are the use cases the Actor's Examples tab is built around. Each is a ready input you can paste into the Apify Console or into RUN_INPUT in the example script.

Track a LinkedIn hashtag: posts, reactions and comments

{ "keywords": ["#semiconductors"], "postedWithin": "week", "maxPostsPerKeyword": 50 }

Every public post under the tag this week, with numLikes, numComments, author and date. Sort by reactions for LinkedIn hashtag analytics: which posts and which authors lead the conversation.

Monitor brand mentions on LinkedIn every day

{ "keywords": ["Your Company", "Your CEO"], "postedWithin": "day", "maxPostsPerKeyword": 100 }

LinkedIn social listening on a schedule. Save this as a task, attach a schedule such as 0 7 * * *, and each morning's dataset is the list of new public posts that mention you.

Search LinkedIn posts by keyword and export to CSV

{ "keywords": ["Jackson Hole Symposium"], "maxPostsPerKeyword": 50 }

A flat row per post: text, authorName, authorUrl, datePosted, numLikes, numComments. The run's Posts view downloads straight to CSV or Excel.

Find the most liked LinkedIn posts on a topic this week

{ "keywords": ["artificial intelligence"], "postedWithin": "week", "maxPostsPerKeyword": 100 }

The viral posts on a topic, ready to sort by numLikes. The top recipe in the example script prints the five most liked.

Search LinkedIn posts by date range for an event

{ "keywords": ["Jackson Hole Symposium"], "startDate": "2026-08-18", "endDate": "2026-08-29", "includeArticles": true }

Every public post and long-form article about an event in its window, each with datePosted, author and text. Useful for research after the fact, when LinkedIn's own search cannot go back by date.

Schedule tip

Any of these inputs becomes a monitor: save it as a task, attach a schedule (0 7 * * * daily, 0 */6 * * * every six hours, 0 9 * * 1 Mondays), and send each run's rows onward with Apify integrations to Slack, Google Sheets, a webhook or a database.

Features

Core Capabilities

• Up to 20 keywords, names or #hashtags per run, each searched as an exact phrase across public LinkedIn posts
• Rolling windows (postedWithin: day, week, month, year) or a custom startDate and endDate
• maxPostsPerKeyword from 1 to 1,000 to cap cost; includeArticles to add long-form LinkedIn articles
• Posts and search pages reused from a shared cache for 7 days, with short windows refreshed every 6 hours (day) and daily (week)
• Datasets download as JSON, CSV or Excel, with a ready-made Posts view
• Works as an MCP tool in Claude, Cursor and ChatGPT (see the install sections below)

Data Quality

• Every post's own timestamp is checked against the window before delivery; posts outside it are never fetched or billed
• matchedKeyword, searchPage and searchRank on every row, so results are traceable
• A post found by two keywords is delivered once
• fromCache and fetchedAt show whether a row came from the shared cache and when it was read
• A keyword or post that cannot be read produces an error row with a plain-language reason, never a charge
• RUN_SUMMARY reports pages fetched and cached, posts delivered and skipped, and why each keyword stopped

It pairs with the rest of the LinkedIn suite: the LinkedIn Posts API for everything one person or company has posted, the LinkedIn Profile API for full profiles of the authors you find (Python example repo), the LinkedIn Company API for the companies that post (Python example repo), and the LinkedIn People Search API to find people by name, title or company.

Coverage, speed and limits

• Coverage. The API finds posts that search engines have indexed from the public side of LinkedIn. Posts from private accounts are not found, and new posts appear within hours, not seconds (three to four hours in testing).
• Speed. A one-keyword, one-page run finishes in under a minute. A busy keyword with hundreds of posts takes several minutes, so the example script starts the run and waits rather than using the synchronous endpoint.
• Comments. Up to 10 per post, which is what LinkedIn shows a signed-out visitor. Reposts and share counts are not shown to guests.
• Not available. Who reacted, private posts, LinkedIn's own relevance ranking, and author filters that need a login.

Pricing

The API is priced per event: one search page event per page of results fetched fresh for a keyword (about 8 to 10 posts per page; pages served from the shared cache are free), and one post scraped event per post delivered. Error rows, posts outside your window and duplicates in a run are never charged. Check the Pricing tab on the Actor page for current rates; the default five-post run here is one page plus five posts. You can set a maximum cost per run in the Apify Console, or pass max_total_charge_usd to start() in Python as the example does.

Usage Examples

Basic Example

The default run in the example script:

{ "keywords": ["Jackson Hole Symposium"], "maxPostsPerKeyword": 5 }

Advanced Example

Two keywords in a custom date range, articles included, using every input parameter:

{

"keywords": ["Jackson Hole Symposium", "#semiconductors"], "startDate": "2026-08-18", "endDate": "2026-08-29", "maxPostsPerKeyword": 100, "includeArticles": true }

From an AI agent over MCP

When an assistant calls the API, keep the cap small and ask a question it can answer from the rows:

{ "keywords": ["#semiconductors"], "postedWithin": "d