DOCX to JSON
Extract Word document structure and metadata as JSON for agents, automation, and retrieval pipelines.
This is structured DOCX extraction for LLM-ready packages, not Word layout reconstruction. Download ZIP includes Markdown, compact Markdown, HTML, JSON, manifest, README, and extracted assets when present.
DOCX to JSON
Upload a DOCX file, convert locally, preview the result, then download the ZIP.
Files stay in the browser. The worker generates a ZIP package locally.
Upload a Word .docx file, then run the matched converter.
Preview panel
Inspect Markdown, HTML, or JSON before downloading.
Convert a supported Office file to inspect Markdown, HTML, and JSON output. After conversion, view full details in the Details workspace.
Details workspace
Opened after conversion for structure summaries, previews, files, and warnings.
Convert a supported Office file to unlock details
After conversion, View details expands structure summaries, Markdown/HTML/JSON preview, package files, and warnings.
What the package includes
The ZIP contains Markdown, compact Markdown, semantic HTML, structured JSON, manifest metadata, extracted images, and a README that documents current extraction limits.
How it works
- Upload a Word .docx file in the browser-local converter.
- Preview the extracted Structured JSON package and inspect conversion warnings.
- Download the ZIP package with Markdown, compact Markdown, HTML, JSON, manifest, and assets.
Best for
- Extract Word document structure for agents, RAG pipelines, and validation jobs.
- Inspect blocks, section paths, styled runs, links, tables, notes, and warnings programmatically.
- Build custom document automation without parsing DOCX XML directly.
Output example
{
"blocks": [
{
"type": "heading",
"level": 1,
"text": "Product requirements"
}
]
}Current DOCX extraction boundaries
The converter extracts document structure, styled runs, links, nested lists, tables, images, basic chart data, text boxes, footnotes, and endnotes. Headers and footers are stored as metadata only. Comments, tracked changes, formulas, macros, embedded objects, and high-fidelity Word layout are not reconstructed.
Privacy boundary
DOCX files are processed in the browser by default. The ZIP package is generated locally before download, and no server upload is required for this converter.
Frequently asked questions
What does DOCX JSON include?
It includes document blocks, section paths, styled runs, links, tables, images, charts, notes, header/footer metadata, and warnings.
Is DOCX to JSON suitable for automation?
Yes. It is designed for agents, RAG ingestion, validation, and custom document workflows.
Are extracted images included in JSON?
Images are exported as assets in the ZIP package and referenced from the JSON and manifest.
Does it include comments or tracked changes?
No. Comments, tracked changes, formulas, macros, and embedded objects are outside the current extraction boundary.