YT-UGC, short for YouTube UGC, is a large-scale user-generated content dataset. The dataset contains 1,500 videos selected from 1.5 million uploaded videos. These videos are divided into 15 categories, and each video has different resolutions and bit rates.

Each video has a MOS score obtained through subjective experiments. For some popular categories, DMOS is added as a reference. In addition to information directly related to video quality, it also contains more than 600 tags related to video content for subsequent research.