AI · other
datasocial/tiktok-5.6B-videos
Public dataset of 5.6 billion TikTok video metadata records in Parquet format
This is a massive dataset hosted on Hugging Face containing metadata for 5.6 billion TikTok videos across roughly 403 GB of Parquet files. It provides detailed fields such as engagement metrics, hashtags, sound identifiers, language, country, and AI-generation labels without requiring custom scraping infrastructure. It is available under a non-commercial CC BY-NC 4.0 license, with commercial access managed separately.
What it's for
- Train and evaluate content recommendation algorithms using large-scale social media metadata
- Analyze viral social media trends and hashtag diffusion across regions and languages
- Study the prevalence and engagement patterns of AI-generated content on video platforms
Features
- 5.6 billion TikTok video metadata records
- Compressed into approximately 403 GB of Parquet files
- Includes engagement metrics, audio IDs, hashtags, and language metadata
- Contains country identifiers and AI-generation labels
- Available under a CC BY-NC 4.0 license for non-commercial research
Pricing
Freemium. Free for non-commercial research (CC BY-NC 4.0); commercial access via datasocial.ai
Details
- Type
- other
- Category
- AI
- Topic
- Datasets
- Open source
- No
- Platforms
- Web
- For
- Machine learning engineers, academic researchers, and data journalists
- Tags
- tiktok, dataset, parquet, machine learning, social media, huggingface
Why it was saved
- @gitatsyndi: Gives researchers, data journalists, and engineers access to detailed social media metadata (including country, language, hashtags, sound IDs, AI-generation labels, and engagement metrics) without needing their own scraper infrastructure.
Related
AI video editor for automated captions, B-roll, and short-form clips
AI tool that turns long videos into short social clips automatically.
AI video generator and editor for short-form faceless clips
AI tool converting text scripts into faceless short-form videos
Turn long videos and podcasts into short, captioned social media clips
Automated video editor applying themed styles, cuts, zooms, and B-roll
Updated