Get post comments

Retrieve comments and the available commenter context for a post.

POST /v1/actions/linkedin-extract-post-commenters/run/live

Authentication and cost

Use an API key in the Bearer header. This action costs 1 credit per returned record. The maximum result count is 100.

Request

bash

curl "$API_URL/v1/actions/linkedin-extract-post-commenters/run/live" \
  -H "Authorization: Bearer $API_KEY" \
  -H "Idempotency-Key: unique-linkedin-extract-post-commenters-001" \
  -H "Content-Type: application/json" \
  -d '{
  "input": {
    "social_id": "urn:li:ugcPost:1234567890123456789"
  },
  "max_results": 10,
  "max_credits": 10
}'

Parameter	Requirement
`input.social_id`	Required non-empty string, at most 500 characters.
`max_results`	Integer from 1 to 100. Default 10.
`max_credits`	Optional positive integer; must cover the requested reservation.
`Idempotency-Key`	Required; 8–128 letters, numbers, underscores, hyphens, or colons.

Output fields

The response contains an output array. Available record fields include:

social_id
full_name
linkedin_profile_handle
linkedin_profile_url
comment_text
comment_time
comment_like_count
comment_reply_count
linkedin_comment_id
linkedin_post_id

Fields not available from the source are omitted. Your parser must tolerate absent properties and empty arrays. Examples are illustrative, not live customer data.

Completion and retries

Both 200 and 202 are valid responses. Read status: a live call can remain queued when capacity is busy. Poll the run until it reaches a terminal state. A terminal failed run can have HTTP 200; inspect its error and status, not just the HTTP code.

Reuse the same idempotency key and request body after a network timeout. A changed request under the same key returns 409.

Coverage notes

Pass social_id returned by profile posts or post search, such as a ugcPost or share URN. The example is illustrative. A displayed activity URL is not proof of the corresponding social ID. Reply counts do not mean every nested reply is included.

See coverage and limits before planning a production dependency. For async execution, replace /run/live with /run/async using the same input.