HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better
cs.CV
Submitted: 2026-07-06
Updated: 2026-09-17
Code: https://github.com/Tencent-Hunyuan/HunyuanOCR
License: http://creativecommons.org/licenses/by-nc-sa/4.0/
The gist: We present HunyuanOCR-1.5, a lightweight end-to-end OCR-specialized vision-language model.
Terminology
Abstract
We present HunyuanOCR-1.5, a lightweight end-to-end OCR-specialized vision-language model. HunyuanOCR unifies document parsing, text spotting, information extraction, text-image translation, and multi-image document understanding within a single end-to-end VLM. Building upon the lightweight architecture of HunyuanOCR-1.0, HunyuanOCR-1.5 does not redesign the backbone, but systematically improves both efficiency and capability. For efficiency, we adapt DFlash to OCR decoding, significantly reducing the latency of long structured outputs such as dense documents, tables, and formulas while preserving output distribution. Powered by DFlash, HunyuanOCR-1.5 achieves a 6.37x Transformer inference speedup and a 2.14x speedup under vLLM, delivering the fastest inference among lightweight OCR VLMs. For capability, we propose Agentic Data Flow, an agent-driven data construction system that transforms model weaknesses into executable data requirements and autonomously performs material search, quality verification, and pipeline development. It substantially improves long-tail capabilities in ancient-script OCR, fine-grained chart and table parsing, multi-image text-centric QA, low-resource multilingual parsing, and document hallucination evaluation. HunyuanOCR-1.5 ranks among the top-tier end-to-end OCR solutions on OmniDocBench v1.6 while achieving new performance milestones across these long-tail tasks. Combined with an upgraded pretraining and post-training recipe, HunyuanOCR-1.5 further extends its capability in high-resolution, long-context, and multi-task scenarios. Experiments demonstrate faster inference, broader OCR capability coverage, and the deployment advantages of a lightweight end-to-end model. We will release the model weights and training code to support future research and real-world OCR applications.
Sources
- Document Parsing Unveiled: Techniques, Challenges, and Prospects for Structured Information Extraction
- Chronicles-OCR: A Cross-Temporal Perception Benchmark for the Evolutionary Trajectory of Chinese Characters
- ChartArena: Benchmarking Chart Parsing across Languages, Scenarios, and Formats
- GPT-4o System Card
- Gemini: A Family of Highly Capable Multimodal Models
- Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
- Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
- Qwen2.5-VL Technical Report
- Qwen3-VL Technical Report
- Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization
- Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
- InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
- HunyuanOCR Technical Report
- MinerU-Diffusion: Rethinking Document OCR as Inverse Rendering via Diffusion Decoding
- DFlash: Block Diffusion for Flash Speculative Decoding
- Fast-dLLM v2: Efficient Block-Diffusion LLM
- AgentInstruct: Toward Generative Teaching with Agentic Flows
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models