Converti Word in Jupyter Notebook online
Students, researchers, and teams draft in Microsoft Word but need a valid .ipynb for JupyterLab, VS Code, or Google Colab. This page is the reverse of our IPYNB to DOCX export: take a .docx hand-in, lecture note, or methodology draft and rebuild notebook cells you can run. Secure server conversion — processed briefly, not stored afterward. Upload a .docx, preview how headings and code styles map to cells, and download nbformat 4.5 JSON with cell IDs included.
We map Word through structured HTML to Markdown, then reuse the same fence-aware pipeline as Markdown to Jupyter so fenced blocks become code cells and narrative stays in Markdown cells. Real Heading 1–6 styles become Markdown # hierarchy; monospace fonts or Word “Code” paragraph styles are preferred signals for executable cells instead of prose. Optional GFM tables keep lab tables inside Markdown cells; optional inline images carry diagrams as data URLs when you need figures in the notebook.
Legacy binary .doc files are not supported—open them in Word or LibreOffice and save as .docx first. Complex floating text boxes and multi-column layouts flatten because notebooks do not have those constructs; keep content linear for best results. We never execute your Python—language tags and style hints only steer kernelspec metadata so Colab and Jupyter open the file cleanly. After download, Run all and fix any fences that Word styled inconsistently.
Use this when a classmate wrote analysis in Word for a course that requires .ipynb, when an instructor distributed .docx notes that should become a lab, or when you edited a notebook in Word via DOCX export and need the interactive format back. Pandoc remains excellent for CLI batches (pandoc file.docx -o file.ipynb); this page is the zero-install path with live preview for one-off Word → notebook conversions without maintaining python-docx scripts on a shared PC. After conversion, skim the preview for misplaced code cells; Word bold-only fake headings without real styles often need a quick fix before you rely on the outline in Jupyter.

