4chan Scraper

Scrape 4chan boards, threads, posts, and images without a login or API key. Give it a board code or a thread URL and get clean structured data: opening posts, every reply, comment text (HTML stripped to plain text), reply and image counts, image and thumbnail URLs, country flags, poster IDs, and timestamps.

What it does

• Board catalog scraper. Give it a board code (for example g, pol, v) and get the opening post of every live thread on that board: subject, comment text, reply count, image count, the OP image, and metadata.
• Full thread scraper. Give it a thread URL and get the opening post plus every reply, flattened into clean one-post-per-row data.
• Board directory. Optionally emit a one-row-per-board list of every available board with its title, page count, and worksafe flag.
• Clean text. Comment HTML is decoded and stripped to readable plain text. Quote references stay readable as >>123.
• Rich media fields. Each post with an attachment carries the full-size image URL, thumbnail URL, original filename, dimensions, file size, and file hash.
• No login required. No account, no password, no API key, no cookies. Works out of the box.

Use cases

• Trend and meme monitoring. Track what is being discussed across boards in near real time for cultural research, brand monitoring, or trend spotting.
• Market and audience research. Pull large samples of posts on a topic to analyze sentiment, language, and emerging interests.
• Dataset building for AI/ML. Collect text and image-link corpora for training, fine-tuning, or NLP experiments.
• Academic and OSINT research. Study discussion dynamics, posting patterns, country distribution, and thread lifecycles at scale.
• Image and media collection. Harvest image URLs, thumbnails, and file hashes from a board or thread for archival or analysis.

Input

Field	Type	Description
boards	array	Board codes to pull the catalog from (the opening post of every live thread). Use the short code with or without slashes, for example g or /g/.
threadUrls	array	Thread links to fetch in full (opening post plus every reply). A short g/123456789 form also works.
listBoards	boolean	When on, also emit a one-row-per-board directory of every available board. Default off.
maxItemsPerSource	integer	Cap on items returned per board or thread. Default 1000.
includeThreadReplies	boolean	When off, thread mode returns only the opening post of each thread URL. Default on.
concurrency	integer	How many sources to process in parallel. Default 3.
proxyConfig	object	Optional proxy configuration. The public read endpoints work without a proxy.

Example input

{
  "boards": ["g", "pol"],
  "threadUrls": ["https://boards.4chan.org/g/thread/123456789"],
  "listBoards": false,
  "maxItemsPerSource": 1000,
  "includeThreadReplies": true,
  "concurrency": 3
}

Output

Each post becomes one clean row. Fields are ordered by importance: identity first, then counts, then content, then author and origin, then media, then metadata.

{
  "type": "post",
  "board": "g",
  "postId": 123456789,
  "threadId": 123456789,
  "isOp": true,
  "url": "https://boards.4chan.org/g/thread/123456789",
  "replyCount": 21,
  "imageCount": 4,
  "subject": "An example thread title",
  "text": "An example post body in plain text.\n>>123456700 quote reference stays readable.",
  "author": "Anonymous",
  "tripcode": null,
  "posterId": "Ab12Cd34",
  "capcode": null,
  "country": "DE",
  "countryName": "Germany",
  "boardFlag": null,
  "boardFlagName": null,
  "image": "https://i.example.com/g/1700000000000.png",
  "thumbnail": "https://i.example.com/g/1700000000000s.jpg",
  "filename": "example.png",
  "fileExt": ".png",
  "fileSizeBytes": 301273,
  "imageWidth": 1280,
  "imageHeight": 720,
  "md5": "zuZHMJMYYp5WY7vM397nWQ==",
  "sticky": false,
  "closed": false,
  "postedAt": "2026-06-01T12:00:00.000Z",
  "time": 1748779200,
  "scrapedAt": "2026-06-18T00:00:00.000Z"
}

Key fields

• postId / threadId / isOp identify the post and whether it is the opening post of its thread.
• replyCount / imageCount are populated on opening posts (the values at scrape time); replies carry null for these by design.
• text is the comment body decoded and stripped to plain text.
• country / countryName and boardFlag / boardFlagName appear only on boards that show flags.
• posterId appears only on boards that show per-thread poster IDs.
• image / thumbnail / filename / md5 and the size fields appear only on posts that include an attachment.

When listBoards is on, board rows carry type (board), board, title, url, worksafe, pages, perPage, and description.

FAQ

Is it free? How is it priced? The actor runs on a simple pay-per-result model. You pay for the rows you receive plus the platform compute the run uses. There is no subscription and no per-run start fee.

Do I need a login, account, or API key? No. It reads public content only, with no account, password, cookies, or API key required.

How many results can I get? A board catalog returns the opening post of every live thread (often 150 or more per board). A full thread can return several hundred posts. Use maxItemsPerSource to cap how many items come back per board or thread, and pass multiple boards or thread URLs to scale up.

How fast is it? Runs are quick. The actor processes sources in parallel (controlled by concurrency) and spaces requests politely. A typical board catalog completes in seconds.

Does it get replies, or just the first post? Board catalog mode returns the opening post of each thread. To get every reply, pass the thread URL in threadUrls with includeThreadReplies on (the default).

Can I get the images? Yes. Each post with an attachment includes the full-size image URL, a thumbnail URL, the original filename, dimensions, file size, and the file hash.

Notes

• Counts (replies, images) reflect the values at scrape time.
• This actor reads public content only.