Data Dumps

Public dataset exports of the archive — posts, comments, polls, and media metadata. Each tarball contains four CSV files (posts.csv, comments.csv, post_images.csv, polls.csv) plus all associated image and video files. Dumps are regenerated nightly. I kindly ask that you don't train generative AI models on this data.

All-Time Dump

Every post and comment in the archive since scraping began.

Download
10,918 posts 17,684 comments 307.1 MB Generated Jun 1, 2026, 3:00 AM EDT

Daily Dump

Posts and comments from the last 24 hours. Replaced each night.

Download
49 posts 77 comments 583.8 KB Generated Jun 1, 2026, 3:30 AM EDT

Manifest

JSON metadata for both dumps: generated times, sizes, and row counts. Useful for checking freshness programmatically.

Download
/dumps/manifest.json

Schema

posts.csv

  • uuid: unique post ID
  • alias: author pseudonym
  • text: post body
  • created_ts: posted at (UTC)
  • upvotes
  • comments_total
  • is_anonymous
  • is_removed
  • link_url: embedded link, if any
  • quote_post_id: FK to posts.uuid

comments.csv

  • uuid: unique comment ID
  • parent_post_id: FK to posts.uuid
  • alias
  • text
  • created_ts (UTC)
  • upvotes
  • is_anonymous
  • is_removed
  • reply_comment_alias: who was replied to

post_images.csv

  • post_id: FK to posts.uuid
  • asset_id: YikYak asset ID
  • media_type: image or video
  • local_path: relative path inside tarball

polls.csv

One row per poll choice.

  • post_id: FK to posts.uuid
  • choice_position
  • choice_text
  • choice_votes
  • view_results_count