2026-08-26 — views
LAION-BVD Releases 10 Million Hour Open Video Dataset for Multimodal Pre-training
Read this because Builders need this scale to train competitive multimodal models across video, audio, and image modalities without relying on proprietary data.
Researchers release LAION-BVD, a dataset containing 80 million videos totaling 10 million hours collected from CommonCrawl.
Dataset Overview
We present LAION-BVD, a large-scale open video dataset for multimodal learning. This collection contains 1.3B platform-specific video URLs collected from CommonCrawl. From these initial URLs, the team downloads 80M videos with a total duration of 10 million hours. The dataset is designed for multimodal pre-training across the video, audio, and image modalities.
Data Processing and Captions
Using content-aware scene detection, the project extracts clips for which they synthetically generate video and audio captions. Models trained on these data achieve competitive performance on standard video-text and audio-text benchmarks. The results show consistent improvements as training or model scale increases.
Image-Text Retrieval
Additionally, the team explores video frames as an alternative source of image-text data by extracting scene-changing frames. These frames exhibit a visual distribution distinct from standard web image corpora. Models trained on this dataset achieve strong image-text retrieval performance.
Practitioner note
Developers should consider using LAION-BVD to significantly expand open access to multimodal videos at an unprecedented scale for their own pre-training pipelines.