Instagram Comments Scraper — $0.50 per 1K

Export public Instagram Reel and video-post comments as structured rows. The Actor is HTTP-only: it does not launch a browser or require an Instagram login. When proxy rotation is enabled, it is explicitly restricted to Apify's DEFAULT non-residential data-center pool.

What you get

• One dataset row per top-level comment or reply.
• Text, timestamp, likes, reply count, author ID/username/name/avatar, and verified status.
• Reel, regular post, and legacy IGTV URL support.
• ID-based de-duplication and exact top-level per-post limits when Instagram exposes enough public rows.
• Public-page hydration first, anonymous Relay cursor pagination, and bounded non-residential session rotation.

Pricing

• Result fee: $0.0005 per dataset row — $0.50 per 1,000 comments or replies.
• A small Actor start event and normal Apify platform/proxy usage may also apply; see the Pricing tab.
• Rows that are not written to the dataset are not charged as results.

Example input

{
  "postUrls": ["https://www.instagram.com/reel/DYSAVzbNlby/"],
  "commentsPerPost": 100,
  "includeReplies": false,
  "repliesPerComment": 20,
  "sortOrder": "popular",
  "proxy": { "useApifyProxy": true },
  "maxConcurrency": 4
}

commentsPerPost=0 and repliesPerComment=0 mean all rows Instagram exposes to the anonymous session. Replies are best effort because Instagram often restricts its public child-comment endpoint. Private, deleted, age/region-restricted, filtered, or login-only comments cannot be returned.

No-residential guarantee

• No Playwright/Puppeteer/browser image is installed.
• Residential proxy groups and custom proxy URLs are ignored by the Actor.
• Proxy-enabled runs use only Apify's DEFAULT non-residential data-center pool.
• Disabling proxy uses direct Apify cloud egress, also non-residential, but may reduce reliability.

Reliability and cost controls

• Cookies and CSRF tokens stay within one short-lived data-center session.
• Short pages, login walls, and soft blocks rotate to a fresh non-residential session and merge unique rows toward the requested limit.
• Opaque cursors are followed sequentially; multiple posts run concurrently.
• HTTP retries, timeouts, page caps, and duplicate-page detection prevent runaway spend.
• A complete run that produces zero valid rows fails instead of reporting a misleading success.

Scrape only public data and comply with applicable laws and platform terms. Avoid collecting or processing personal data without a lawful purpose.