Research · Oct 30, 2026, 00:00 UTC

Youtu-Parsing-Omni parses documents, images, audio and video in one 5B model

A single 5 billion parameter checkpoint reports state-of-the-art results on OmniDocBench v1.6 (96.96 Overall) and leads open-weight models on OmniParsingBench (75.08 Avg.). Code and weights are publicly available.

• ModelYoutu-Parsing-Omni
• Topicresearch
• Topicopen-weights
• Topicbenchmark