X Twitter Replies Scraper

Extract public replies from X/Twitter tweet URLs or numeric tweet IDs. The actor is provider-backed, requires no cookies, and returns clean reply records for sentiment analysis, conversation monitoring, social listening, creator research, support triage, and AI-agent workflows.

What it does

This actor scrapes the public reply timeline for specific X/Twitter posts. It accepts full tweet URLs or tweet IDs, paginates through reply results, dedupes repeated replies, and saves one normalized dataset row per reply.

Use it for:

• Brand and campaign reply monitoring
• Sentiment analysis on public tweet conversations
• Customer support and escalation discovery
• Creator, influencer, and community research
• Lead and intent discovery from public replies
• MCP optimized AI-agent research pipelines

Do not use it for keyword search, profile timelines, quote tweets, retweeters, private accounts, deleted tweets, DMs, or login-only data.

Input

Important fields:

• tweetUrls - one or more X/Twitter tweet URLs or numeric tweet IDs
• rankingMode - Relevance, Recency, or Likes
• maxReplies - total reply rows to save across all tweets
• maxPagesPerTweet - pagination depth per tweet
• startCursor - optional continuation cursor from RUN_SUMMARY.nextCursors
• includeRaw - include raw provider payloads for debugging
• dedupeReplies - skip duplicate reply tweets across pages

Example:

{
  "tweetUrls": ["https://x.com/OpenAI/status/2065225362544726371"],
  "rankingMode": "Relevance",
  "maxReplies": 100,
  "maxPagesPerTweet": 10,
  "dedupeReplies": true
}

Output

Each dataset item is a flat, AI-agent-friendly reply record:

• parentTweetId, parentTweetUrl, rankingMode
• replyId, replyUrl, text, createdAt, language
• conversationId, inReplyToTweetId
• authorUsername, authorName, authorId, authorUrl
• authorVerified, authorFollowers, authorFollowing, authorBio, authorProfileImageUrl
• replyCount, retweetCount, quoteCount, likeCount, viewCount, bookmarkCount
• hashtags, mentions, urls, mediaUrls, source
• provider, providerPage, scrapedAt

The actor also writes:

• OUTPUT - concise machine-readable run summary
• RUN_SUMMARY - detailed telemetry, warnings, provider status codes, and next cursors

Public data only

This actor collects public reply data visible through the provider-backed X/Twitter reply endpoint. It does not unlock private accounts, hidden replies that are not exposed upstream, deleted posts, DMs, private follower graphs, or login-only fields.

Pricing

The actor uses pay-per-event pricing:

• Actor start: $0.00005
• X/Twitter reply saved: $0.008

Reliability

The actor uses provider retries, cursor pagination, status-code telemetry, duplicate detection, cost-cap status messages, clean empty-result handling, and graceful terminal summaries. Missing provider credentials or upstream failures are written to OUTPUT and RUN_SUMMARY with actionable status messages.

Tips

• Use Relevance for the default conversation view.
• Use Recency for fresh monitoring.
• Use Likes when you want high-engagement replies first.
• Use startCursor only when continuing the same tweet and ranking mode from a prior run.
• Keep includeRaw disabled unless debugging field drift.