💡How large are Thordata's social media datasets?
Officially 20 billion YouTube comments, 90 billion TikTok metadata records, and 1 billion X (formerly Twitter) posts — suited to sentiment modeling, trend forecasting, and expanding LLM corpora; not for creator-dashboard metrics, deleted content, or user-privacy attributes. Collected from public pages; confirm fields and delivery format on the official site before requesting a sample.
PreviousAre low-resource ASR corpora available off-the-shelf?
NextDoes Thordata offer multilingual OCR datasets?
Last updated
Was this helpful?