Youtu-Parsing: Perception, Structuring and Recognition via High-Parallelism Decoding

2026-01-28

AI Summary

This work addresses the robustness and efficiency bottlenecks in multimodal content extraction—encompassing text, tables, formulas, and figures—within document intelligence, particularly under challenging conditions such as multilingual settings, handwritten inputs, and rare characters. To this end, we propose a decoupled and feature-reusable document parsing framework that integrates a dynamic-resolution Vision Transformer visual encoder with a prompt-guided Youtu-LLM-2B large language model. Our approach introduces a novel dual high-parallel decoding mechanism combining token-level and query-level parallelism, achieving 5–11× and 2× speedups respectively while maintaining output quality. The method attains state-of-the-art performance on both OmniDocBench and olmOCR-bench, significantly enhancing structured document parsing efficiency and cross-domain generalization capability.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This paper presents Youtu-Parsing, an efficient and versatile document parsing model designed for high-performance content extraction. The architecture employs a native Vision Transformer (ViT) featuring a dynamic-resolution visual encoder to extract shared document features, coupled with a prompt-guided Youtu-LLM-2B language model for layout analysis and region-prompted decoding. Leveraging this decoupled and feature-reusable framework, we introduce a high-parallelism decoding strategy comprising two core components: token parallelism and query parallelism. The token parallelism strategy concurrently generates up to 64 candidate tokens per inference step, which are subsequently validated through a verification mechanism. This approach yields a 5--11x speedup over traditional autoregressive decoding and is particularly well-suited for highly structured scenarios, such as table recognition. To further exploit the advantages of region-prompted decoding, the query parallelism strategy enables simultaneous content prediction for multiple bounding boxes (up to five), providing an additional 2x acceleration while maintaining output quality equivalent to standard decoding. Youtu-Parsing encompasses a diverse range of document elements, including text, formulas, tables, charts, seals, and hierarchical structures. Furthermore, the model exhibits strong robustness when handling rare characters, multilingual text, and handwritten content. Extensive evaluations demonstrate that Youtu-Parsing achieves state-of-the-art (SOTA) performance on both the OmniDocBench and olmOCR-bench benchmarks. Overall, Youtu-Parsing demonstrates significant experimental value and practical utility for large-scale document intelligence applications.

Problem

• document parsing
• content extraction
• structured document understanding
• multilingual recognition
• handwritten text

Innovation

• high-parallelism decoding
• token parallelism
• query parallelism
• region-prompted decoding
• document parsing