Provides 1.3 billion platform-specific video URLs extracted from CommonCrawl along with crawl metadata (no media included), serving as the source corpus for the LAION-BVD multimodal video dataset; distributed on Hugging Face in Parquet format.