TikTok Metadata Dump Hits Hugging Face: 5.6B Rows, Zero Tickers, All Parquet

A developer using the handle hashfunction has published metadata for roughly 5.6 billion public TikTok videos on the open-source AI repository Hugging Face, available at no cost under the datasocial account. The dataset spans July 2014 through October 2026 and consists of 460 GB of Parquet files split by month, with each row containing captions, hashtags, sound IDs, on-screen text, TikTok Shop product and seller IDs, and counts for views, likes, comments, shares, saves, and downloads. Hugging Face's counter shows 1,181 downloads since release, and the files are also hosted online at datasocial.ai.

The release does not include creator handles, but it does carry TikTok's own flags for whether a video is labeled AI-generated and whether TikTok has kept it off the For You page. The AI-generated label is often blank for older items, which the publisher attributes to the underlying archive. The data is licensed CC BY-NC 4.0 and is offered free for research and personal use, while commercial access, creator profiles, daily updates, and the scraper's source code are routed through datasocial.ai.

A write-up on datasocial.ai describes how the collection was assembled: generated device identities designed to pass as Android phones, reverse-engineered request signatures, and a spoofed TLS handshake used to reach TikTok's private mobile API without any login. According to the write-up, the system gathered 3.23 billion creator profiles, 5.94 billion videos, and 2.8 billion comments over three weeks.

The publication lands as legal pressure on data scraping intensifies. In October 2025, Reddit filed suit against Perplexity and three data-scraping firms, alleging an "industrial-scale" scheme to harvest its content for AI training. In July, a federal judge largely declined to dismiss the case, according to Law360, with surviving claims including DMCA allegations that SerpApi circumvented Google's anti-bot protections. TikTok is not a party to that litigation, which involves scraping through Google search results rather than TikTok's mobile API, and Section 3.4 of TikTok's U.S. terms of service bars extracting data from the platform with automated software without prior written approval.