Douyin Video Comments API

Douyin Video Comments

Features

• Direct video targetingResolve the numeric video ID from one public douyin.com or iesdouyin.com URL whose path contains /video/{numeric_id}.
• Bounded one-page collectionRequest up to 20 comment rows from the first source page, with results_limit controlling the stored subset from 1 through 20.
• Canonical comment contextReturn CommentItem 1.0 records with text, author identity, video and parent relationships, publication and collection times, and visible like or reply counts.

Parameters

How to Use

Set a direct public Douyin video URL, submit the asynchronous task, and read the normalized comments stored for that task.

1. Set DOUYIN_VIDEO_URL to a public douyin.com or iesdouyin.com URL with a /video/{numeric_id} path.
2. POST url and the optional results_limit to /v1/douyin/video-comments, then save task_id.
3. Poll /v1/tasks/{task_id} until the task succeeds or fails.
4. Read data.results.items and follow next_cursor only while has_more is true to continue through records already stored by SocQ.

curl -X POST "https://api.socq.ai/v1/douyin/video-comments" \
  -H "Authorization: Bearer $SOCQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"url\":\"${DOUYIN_VIDEO_URL}\",\"results_limit\":20}"

# Poll GET /v1/tasks/{task_id}
# Read data.results.items and follow next_cursor while has_more is true

Best Use Cases

• Audience-response review: Use comment text, author identity, and visible like or reply counts to review responses to a known public video.
• Comment relationship mapping: Connect returned comments to the source video and any supplied parent reference with post_id and parent_id.
• Engagement snapshot comparison: Compare visible comment signals across selected videos or collection times while retaining collected_at as the observation timestamp.

Pro Tips

• Resolve a shared or redirected link to its final accepted /video/{numeric_id} URL before submitting it, so host and path validation can succeed.
• Omit results_limit to request 20 rows, or choose a smaller integer when only a subset of the one returned source page is needed.
• Reuse the original task_id and follow next_cursor only when has_more is true; that cursor reads stored rows and does not fetch another source page.
• Use the comment id as the primary deduplication key, and allow parent_id or post_id to remain null when the source row omits a relationship.
• Treat author details, published_at, and individual metrics as optional, and use collected_at to timestamp every result snapshot.
• Escape comment text before inserting it into HTML because public comments are user-generated content.

Related APIs