---
base_model: Qwen/Qwen3-VL-4B-Instruct
library_name: transformers
pipeline_tag: image-text-to-text
tags:
  - qwen3-vl
  - document-vqa
  - agentic-ocr
---

# AgenticOCR-4B

Qwen3-VL-4B-Instruct model trained for the [AgenticOCR](https://github.com/opendatalab/AgenticOCR) project: query-driven, on-demand visual evidence extraction for document RAG.

This repository contains the model weights, configuration, tokenizer, chat template, and visual preprocessor files required for inference.