Tencent releases WeVisDoc-4B, an end-to-end document parser

Tencent has released WeVisDoc-4B, an end-to-end document parser for page images that converts pages into structured Markdown with LaTeX formulas and HTML tables. Fine-tuned from Qwen3-VL-4B-Instruct, the model achieves an Overall score of 95.38 on OmniDocBench v1.6 and ranks first among compared parsers across all four reported settings.

• WeVisDoc-4B is fine-tuned from Qwen3-VL-4B-Instruct.
• It outputs structured Markdown, LaTeX formulas, and HTML tables.
• The model achieves an Overall score of 95.38 on OmniDocBench v1.6.
• It leads compared end-to-end parsers across all four reported settings.

The release includes installation instructions for vLLM and client dependencies, along with examples for local inference and serving the service over a network.