LinkedIn Get Direct Messages Actor

Apify actor that retrieves all direct messages from your LinkedIn messaging inbox with comprehensive conversation details, participant information, and message metadata.

Features

• Retrieves all conversations from LinkedIn messaging inbox
• Extracts detailed information for each conversation:
  • Thread URL
  • Participant LinkedIn profile URL
  • Last message content
  • Sender information (firstname, lastname, occupation, company)
  • Message direction (from me or from contact)
  • Read/unread status
  • Message timestamp
• Uses LinkedIn session cookies for authentication
• Follows the same authentication pattern as other Apify LinkedIn actors
• Multi-selector strategy for robust element detection
• Automatic scrolling to load all conversations

Input

The actor requires the following input parameters:

Required

• liAtCookie (string, required): Your LinkedIn li_at cookie value for authentication
  • Example: "AQEDATEtZ..."
  • How to get it: Open LinkedIn in your browser, open Developer Tools (F12), go to Application/Storage → Cookies → https://www.linkedin.com, and copy the value of the li_at cookie

Optional

• jsessionid (string, optional): Your LinkedIn JSESSIONID cookie value for additional authentication

  • Example: "ajax:1234567890123456789"
  • Found in the same location as li_at cookie

• userAgent (string, optional): Custom user agent string

  • Default: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36"
  • Use a recent Chrome/Edge user agent for best results

Input Example

{
    "liAtCookie": "AQEDATEtZ...",
    "jsessionid": "ajax:1234567890123456789",
    "userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36"
}

Output

The actor outputs a single dataset item containing:

• success (boolean): Whether the operation was successful
• message (string): Status message
• conversationCount (number): Total number of conversations retrieved
• conversations (array): Array of conversation objects
• timestamp (string): ISO timestamp of when the data was retrieved

Conversation Object Structure

Each conversation in the conversations array contains:

• threadUrl (string): Direct URL to the conversation thread
• linkedInUrl (string): LinkedIn profile URL of the contact
• message (string): Content of the last message in the thread
• lastMessageFromUrl (string): Profile URL of the last message sender
• firstname (string): First name of the contact
• lastname (string): Last name of the contact
• occupation (string): Current job title/occupation of the contact
• companyName (string): Company name where the contact works
• isLastMessageFromMe (boolean): Whether the last message was sent by you
• readStatus (boolean): Whether the conversation has been read (true = read, false = unread)
• lastMessageDate (string): Human-readable date of the last message (e.g., "2h", "Yesterday", "Sep 01")
• timestamp (string): ISO timestamp of the last message (if available)

Output Example

{
    "success": true,
    "message": "Successfully retrieved 25 conversations",
    "conversationCount": 25,
    "conversations": [
        {
            "threadUrl": "https://www.linkedin.com/messaging/thread/2-MGFjMTExN2YtOWI3NS00NmJlLWEyNjMtNjhlZDFjZWZkNDVkXzEwMA==/",
            "linkedInUrl": "https://www.linkedin.com/in/ACoAABJnny0ButYTqma7Tmq-SqQDoFJVPpqBIr8",
            "message": "Salut Pierre, j'ai vu que tu as jeté un œil à Hootsuite...",
            "lastMessageFromUrl": "https://www.linkedin.com/in/ACoAABJnny0ButYTqma7Tmq-SqQDoFJVPpqBIr8",
            "firstname": "Charlotte",
            "lastname": "Clemens",
            "occupation": "VP of Sales Enterprise Europe - Hootsuite & Talkwalker",
            "companyName": "Hootsuite",
            "isLastMessageFromMe": false,
            "readStatus": true,
            "lastMessageDate": "Sep 01",
            "timestamp": "2024-09-01T10:30:00.000Z"
        }
    ],
    "timestamp": "2025-01-15T14:23:45.123Z"
}

How It Works

1. Authentication: Sets LinkedIn session cookies (li_at and optionally JSESSIONID)
2. Navigation: Opens LinkedIn messaging page (https://www.linkedin.com/messaging/)
3. Conversation Loading: Automatically scrolls to load all available conversations
4. Data Extraction: For each conversation:
   • Extracts thread URL, message content, timestamp, and read status from the messaging list
   • Navigates to the contact's profile page
   • Extracts detailed profile information (name, occupation, company)
5. Output: Returns all conversation data in a structured format

Performance

• Execution Time: ~2-5 seconds per conversation (due to profile information extraction)
• Scrolling: Automatically scrolls up to 5 times to load all conversations
• Anti-Detection: Includes user-agent spoofing and webdriver masking

Error Handling

The actor will throw an error if:

• The liAtCookie parameter is missing
• LinkedIn returns a non-200 status code
• The messaging page fails to load
• Authentication fails (invalid cookies)

Notes

• Cookie Expiration: LinkedIn cookies expire periodically. If the actor fails, try getting fresh cookies from your browser.
• Rate Limiting: LinkedIn may rate-limit requests if you run the actor too frequently. Use reasonable delays between runs.
• Profile Privacy: Some profile information may not be available if the contact has restricted their profile visibility.
• Headless Mode: The actor runs in headless mode by default for better performance.

Security

• Never share your li_at cookie publicly as it grants full access to your LinkedIn account
• Use Apify's secret input feature to store cookies securely
• Cookies are only used for authentication and are not stored or logged

Use Cases

• Export LinkedIn conversations for backup
• Analyze messaging patterns and response rates
• Integrate LinkedIn messaging data with CRM systems
• Monitor unread messages and follow-ups
• Generate reports on networking activities

Related Actors

• linkedin-send-dm: Send direct messages to LinkedIn profiles
• linkedin-like-post: Automatically like LinkedIn posts
• linkedin-comment-post: Automatically comment on LinkedIn posts
• linkedin-quote-post: Repost LinkedIn content with your thoughts

Support

For issues or questions, please contact the development team.

Version History

• 1.0.0 (2025-01-15): Initial release
  • Basic conversation retrieval
  • Profile information extraction
  • Read status detection
  • Multi-selector strategies for robust element detection