X.com (Twitter) Tweet Replies Scraper

Paste any tweet. Get every reply back, from every author, walked all the way down the conversation tree. Clean, structured data straight to your spreadsheet, CRM, or dashboard.

💬 What it does

Give it any tweet on X (formerly Twitter) and this tool returns every reply we can see:

• From every author. Not just the tweet owner, not just the first few. Every public reply.
• Replies of replies. Walks the whole conversation tree, up to 5 levels deep.
• With a parent link on every row so you can rebuild the full thread structure downstream.

For every reply you get:

• The full text (no 280-character truncation on long-form notes)
• Author handle, display name, avatar, verification badges
• Likes, retweets, replies, quotes, bookmarks, view count
• Attached photos and videos (turn on the download add-on to save the files)
• Hashtags, mentions, and expanded URLs
• parentReplyId and depth so the reply tree is reconstructable, plus streaming delivery position
• Precise creation timestamp in ISO 8601

📋 How to use it

Step 1: Paste your tweets

• Tweet URLs or IDs: one per line. Full URLs, share links, or plain tweet ID numbers all work (x.com, twitter.com, mobile.twitter.com, fxtwitter.com, vxtwitter.com).
• Up to 500 tweets per run.

Step 2: Set the walk depth and cap

• Max replies per tweet: user-controlled limit per input tweet. Default 100; no product-level maximum.
• Max depth: how deep down the reply tree we walk. 1 = direct replies only, 2 = replies plus reply-to-replies (default), up to 5 for full deep-dive.
• Sort: Top (engagement-weighted, default), Latest, Oldest, or Most Liked. Sorting applies inside each streamed conversation buffer; position is assigned before concurrent storage, so physical dataset order can differ.

Step 3: (Optional) Filter your results

Every filter is off by default. Filters are free. Replies dropped by a filter are never billed.

• Minimum likes / retweets / replies: drop low-engagement noise.
• Verified authors only: skip non-verified handles.
• Only replies with media: keep photos and videos.
• Exclude the OP's self-replies: drop the original tweet author's own replies (use our Thread Scraper for those).
• Keyword include / exclude regex: case-insensitive regex match on reply text. Great for filtering out spam like airdrop|crypto|giveaway.

Step 4: (Optional) Add-ons

Every add-on is off by default. Each is billed only when it returns a real result. Failed lookups cost nothing.

• Download photos and videos: save every image and video into your run's key-value store. Each media item gets a signed download link (kvUrl). Billed once per reply that had at least one file saved.
• Author enrichment: for every unique reply author, fetch the full profile (followers, bio, join date, verified badge). Billed once per unique handle across the run, cached across every input tweet.

Step 5: Run

Click Start. The root row is written immediately. Each reply enters the delivery queue when its concurrent detail fetch finishes. Base replies are pushed one row per combined dataset/charge call with storage concurrency 4; the queue starts work at 8 replies or 100 ms. Add-on rows remain serialized because their independent charges require stricter settlement handling.

Each streamed root, reply, and summary row carries an internal _deliveryId. Customer views omit it; restart recovery scans it to avoid duplicate output or charges.

📊 What you get

Every row is one reply. Four tabs group the fields by what you care about:

Tab	What is in it
📋 Overview	OP handle, position, depth, reply text, top engagement counts, posted date, reply URL
💬 Replies	Full spread: reply ID, parent ID, in-reply-to fields, every engagement count, URL
👤 Authors	Author profile columns (populated when the Author enrichment add-on is on)
🖼️ Media	Replies that have media. Signed download link per file when the Download add-on is on

Export from the Output tab as JSON, CSV, or Excel.

Every row has a status field. Success means that reply row is complete. Replies stream page by page and remain valid if a later page fails; the final summary row reports partial_replies when conversation coverage is incomplete. Summary and error-only rows are free.

💰 Pricing

You only pay for replies you actually keep. Every event is billed per result.

Event	Rate	When it fires
Reply scraped	$0.50 / 1,000	Once per surviving reply row
Media extracted	$2.00 / 1,000	Once per reply with at least one file saved
Author enriched	$1.00 / 1,000	Once per unique reply author successfully resolved

Deleted tweets, private accounts, and rate-limit errors are always free. Replies dropped by a filter are always free.

💡 Good to know

• Every reply, from everyone. Not just the first few. Not just from the tweet author. Every public reply we can see.
• Filters that save you money. Drop low-quality replies before they cost you anything.
• Streaming survives slow threads. Flat reply rows are saved after each conversation level. A later timeout does not erase earlier delivered replies.
• Clean data ready for a spreadsheet. Numbers as numbers, dates as dates, links as links. No cleanup needed.
• Author enrichment is cached across the whole run. The same handle across many input tweets bills once.
• Long text is never truncated. Long-form note tweets come back in full.
• URL forms accepted: x.com/user/status/123, twitter.com/user/status/123, mobile.x.com/..., share links, and bare numeric IDs all work.
• Media files are stored in your Apify key-value store. How long they are kept depends on your Apify plan.

💻 For developers

Call this actor through the Apify API:

{
    "tweetUrls": ["https://x.com/karpathy/status/1615398117683388417"],
    "maxRepliesPerTweet": 100,
    "maxDepth": 3,
    "sort": "top",
    "minLikes": 5,
    "verifiedOnly": false,
    "hasMediaOnly": false,
    "excludeOpSelfReplies": true,
    "keywordExcludeRegex": "airdrop|crypto|giveaway",
    "outputMode": "flat",
    "enrichAuthors": true,
    "downloadMedia": false
}

Only tweetUrls is required. Every other field is optional.

❓ FAQ

Do I need a Twitter or X account? No. Nothing to sign up for.

Which links work? Anything you can copy from X or Twitter. Full URLs, share links, or plain tweet ID numbers. Works with x.com, twitter.com, mobile.twitter.com, fxtwitter.com, vxtwitter.com.

How many replies will I actually get? Depends on the tweet. The default cap is 100 replies per input tweet. You can set any positive safe-integer limit; there is no separate product-level maximum. Some replies are hidden by X and cannot be seen without logging in, so you will not always get 100% of what the reply count on the tweet shows.

Can I filter out spam and giveaway bots? Yes. Turn on "Verified authors only" or add a regex like giveaway to the exclude filter. Set a minimum like count too. Anything the filter drops does not bill.

What if the tweet was deleted? You will see a clear status message in the results telling you what happened. That run is free.

Can I get every reply's author profile? Yes. Turn on the Author enrichment add-on. You get followers, bio, join date, and the verified badge for every unique reply author. Cached across the whole run so you pay once per unique handle.

Can I download the images and videos in replies? Yes. Turn on the Download media add-on. Files land in your Apify key-value store, ready to grab.

What does depth mean? 0 is the original tweet. 1 is a direct reply. 2 is a reply-to-a-reply. And so on, up to your Max depth setting.

How fast is it? About 100 replies per minute at the default memory tier. Bigger memory tiers run faster.

Can I use my own X cookies? No. This actor does not accept user cookies.

🔗 Other tools you might like

• X Thread Scraper: The tweet author's own reply chain, unrolled as a full thread with a free Markdown export.
• X Tweet Scraper: Search tweets by keyword, hashtag, or handle.
• X Profile Scraper: Every tweet from any public profile.

⚖️ Disclaimer

This tool is intended for personal, research, and educational use. You are responsible for complying with X.com's Terms of Service and applicable laws in your jurisdiction. The developer is not liable for misuse. Data availability depends on what X shows publicly at the moment you run it.

Keywords: x replies scraper, twitter reply scraper, twitter comments scraper, x reply tree, twitter reply tree, x conversation scraper, tweet reply extractor, twitter reply analytics, twitter reply monitoring, x comment mining, apify twitter, twitter replies to csv, twitter reply download, tweet response scraper, twitter comment extractor