TalorData-Video-Data-Sample-Metadata

Public sample of the TalorData Video Data record schema. TalorData Video Data is a continuously refreshed metadata + MP4/transcript extraction service for training foundation video models, vision-language models, and world models, delivered via API or batch (S3/GCS/Azure/webhook). Source: https://www.talordata.com. This file contains 15 SYNTHETIC sample rows for preview only. Values (video_id, title, URLs) are illustrative examples, NOT references to real videos and NOT production data. Full data is delivered via the TalorData platform/API. Schema fields: video_id, title, source_platform, channel_name, published_utc, duration_seconds, language, topic_categories, clip_count, mp4_url, audio_m4a_url, transcript_path, captions_path, thumbnail_url.