Convert HWP to Markdown in your browser
Convert HWP 5.0 and HWPX to Markdown without uploading the document. HanOpen writes readable text, Markdown markup and, where needed, HTML table blocks; the default picture option packages the Markdown with an assets folder when supported pictures are extracted. Without extracted pictures, that setting produces an MD file.
Your file never leaves this device.
Drop files here
Accepted: .hwp, .hwpx. Several files at once are fine.
Up to 200 MB per file on this device.
Files (0)
No files yet. Drop one above or choose from your device.
Waiting for a file.
The download button appears here once a conversion finishes.
Conversion summary
After a conversion this card lists what was detected and what was written to the output.
Layout is not preserved. Tables with merged cells are written as HTML tables inside the Markdown, and headers and footers are left out.
How to use it
- Add the HWP or HWPX documents you want to turn into notes, then review the selected file names.
- Choose whether pictures belong in an assets folder, inside the Markdown, or should be omitted, and start the conversion.
- Save the single result or the batch ZIP, extract any archive, and preview the Markdown together with its picture folder before moving it elsewhere.
What is preserved
Simple rectangular tables can use GFM pipe syntax. Merged cells, nested tables and cells containing multiple paragraphs require HTML table blocks instead. Footnote references use Markdown notation, while page headers and footers are omitted from this output. Detected and emitted counts describe individual elements, not a guarantee that the destination editor will display them identically.
| HWP element | In this output |
|---|---|
| Paragraphs and text | Kept as structure and formatting |
| Character formatting | Kept with simplified formatting |
| Tables | Kept with simplified formatting |
| Merged cells | Kept with simplified formatting |
| Text boxes and shapes | Kept with simplified formatting |
| Images | Kept with simplified formatting |
| Footnotes | Kept as structure and formatting |
| Endnotes | Kept as structure and formatting |
| Headers | Not included; a notice is shown |
| Footers | Not included; a notice is shown |
| Numbering and bullets | Kept with simplified formatting |
| Equations | Kept as plain text |
| OLE objects | Replaced with a placeholder |
| Charts | Replaced with a placeholder |
| Distribution-only documents | Not included; a notice is shown |
| Password-protected documents | Not included; a notice is shown |
This guide describes what the tool handles. Review the result and any warnings before using your output.
About the formats
Markdown stores text and formatting marks in an ordinary text file. It is useful when the next step involves revising prose, reviewing changes or organising source material. A heading mark expresses a text level, not a page position. List markers express sequence, not the original indentation measurements. HWP and HWPX contain richer document structures, so this conversion deliberately reduces some presentation details. The result may mix Markdown with HTML because pipe tables have no syntax for merged cells. Picture references remain separate from the prose unless you select embedded images. Whether a destination accepts HTML tables, data URIs or footnote notation is a separate question: try a small document in that destination before processing a large collection.
Frequently asked questions
- Why did I receive a ZIP rather than an MD file?
- With the default image setting, successfully extracted pictures and the Markdown travel together in one archive. If there are no extracted pictures, the result is an MD file instead. Unpack it before previewing the document, and retain the relationship between the MD file and its assets directory. A batch download may also be an outer ZIP containing several individual results.
- Why are there HTML tags around a table?
- Pipe tables cannot represent every HWP table. The HTML form can express row and column spans as well as richer cell contents. Those tags are part of the result, not an error message. A destination that removes embedded HTML may hide or simplify the table.
- Which image setting works best for text extraction?
- For prose review or a text prompt, omitting pictures avoids long encoded image strings. The assets option keeps recoverable image files beside the document. Embedding pictures makes a single MD file but may enlarge it substantially, and some Markdown viewers ignore data URI images.
- Will mathematical formulas become editable equations?
- The current output preserves a formula script as text where the parser exposes it. It does not translate that script into a mathematical notation understood by every Markdown renderer. Compare important symbols with the source and rewrite the expression in your destination’s supported syntax when necessary.
- Can the converter read a protected HWP attachment?
- Password protection, distribution restrictions, DRM and encrypted HWPX bodies stop conversion. Changing the filename extension does not remove those restrictions. Request an unrestricted source document or another export from the sender if you are allowed to use its contents.
- Is the generated Markdown ready for a language model?
- It is editable source text, but preparation still matters. Remove irrelevant boilerplate, check tables, and decide whether images or footnotes contain essential evidence. Text extraction does not perform OCR on picture contents, verify statements, or decide whether you should send the resulting text to another service.
Keep picture paths intact
Relative picture paths start from the location of the Markdown file. Moving only the MD file can therefore leave broken images even though conversion succeeded. Move its assets folder with it, or update the paths when you intentionally reorganise the material. Renaming an image also requires updating the corresponding reference. After extracting a ZIP, open one picture directly to distinguish a missing file from a destination that does not render the reference. Do not infer that every picture survived solely from the presence of an assets folder; inspect the conversion summary for image warnings.
Review a table as information
Before publishing a converted table, compare its headings, merged labels, units and explanatory notes with the source. An HTML block may preserve a merged region while a paste operation elsewhere flattens it. In a constructed example, a timetable has one date cell spanning three sessions: after pasting, verify that each session still belongs to that date. If the destination accepts only pipe tables, decide explicitly whether to repeat the date in each row or explain the grouping above the table. That editorial decision changes the representation and should be checked rather than applied blindly.
Build a useful document note
A sensible note keeps the source filename, document date and the relevant section title near an excerpt. These are choices for your own note, not fields that the converter invents. In a constructed meeting example, retain the decision paragraph and its supporting footnote before summarising the discussion. Check that a numbered item has not become separated from its explanation. If a title depended only on a large font instead of a heading structure, promote it manually after reading the surrounding paragraphs. Conversion provides editable material; it does not choose a reliable outline for you.
Distinguish a warning from lost prose
A summary can report an unsupported shape, formula representation or page feature even when the visible paragraphs remain useful. Read the warning against the part of the document you need. Conversely, a successful download only confirms that an output file was created. Search for a distinctive sentence near the end, compare an important table and inspect any referenced figure. The current text corpus gate checks 41 supported sample documents against text extracted by the same parsing model; it does not establish compatibility with every Markdown editor or independently prove every original layout feature.
Other options
Keep the HWP or HWPX source alongside your notes when you need an audit trail. Choose HTML for a standalone reading copy, or DOCX when the next task is editing a document with page and paragraph formatting. Markdown is the practical choice here when the text itself will be revised and reorganised.