PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training
PaddleOCR-VL-1.6 applies region-aware data optimization and progressive reinforcement learning post-training to achieve 96.33% on OmniDocBench v1.6.
Published Jun 2, 2026▲ 26 on Hugging FaceCode ★ 90,719arXiv ↗
Only vote on papers you've read. Sign in with GitHub to vote.
PaddleOCR-VL-1.6 delivers a precise region-aware optimization and progressive post-training recipe that achieves a state-of-the-art 96.33% benchmark score, though its lack of ablation and variance reporting leaves open whether the gains reflect genuine mechanism or targeted overfitting.
Abstract
We introduce PaddleOCR-VL-1.6, an upgraded compact document parsing model built upon PaddleOCR-VL-1.5. Although PaddleOCR-VL-1.5 establishes a strong 0.9B baseline, its remaining errors concentrate in under-optimized regions where model behavior is unstable, data coverage is sparse, or supervision is unreliable. Rather than expanding the training corpus indiscriminately, PaddleOCR-VL-1.6 introduces a region-aware data optimization framework that identifies weak regions from the previous model, applies targeted enhancement to these regions, and improves the reliability of supervision signals. It further adopts a progressive post-training recipe based on curated data selection and reinforcement learning, pushing model performance to a higher level through staged optimization. PaddleOCR-VL-1.6 achieves a new state-of-the-art score of 96.33% on OmniDocBench v1.6, demonstrates strong competitiveness against top-tier VLMs, and provides a practical post-training recipe for the PaddleOCR-VL series.