DOCX merger

• Choose two to six DOCX files based on compatible global templates.
• Keep all for displayed file order, or enter every file number once, such as 2,1.
• Merge the documents and download merged.docx. Open it in your editor to review pagination and layout.
• Cancel or Reset terminates processing and clears the result.

Join portrait and landscape reports with different Normal and custom styles, independent lists, tables, images and section headers.

• The global docDefaults, theme, settings, fontTable and webSettings resources must match. Different named/default paragraph and table styles are remapped; paragraphs and tables without explicit styles require declared source defaults. Missing run styles remain absent, retaining the source style cascade.
• Each appended document begins on a new page. An explicit source logical page start is retained; otherwise it restarts at one to preserve odd/even headers. Internal section properties are retained. No rendering, field updates or universal layout-fidelity claim. A table-ending section may need a minimal closing paragraph.
• Two to six inputs; 20 MiB compressed combined, 1,000 total ZIP entries, 16 MiB per entry and 64 MiB expanded combined. UTF-8 transitional DOCX only. Up to 50 combined explicit sections and 2,000 styles/numbering definitions per input.
• XML limits: 8 MiB per part, 16 MiB combined, 200,000 events per part and 400,000 combined elements/attributes/text; 64 XML levels, 128 attributes and 64 style-inheritance levels. Copied XML work is additionally capped at 400,000 events. Resource relationships are limited to 32 levels. A 60-second local worker deadline applies.
• Output is limited to 32 MiB compressed, 64 MiB expanded, 1,000 parts and 16 MiB XML. Limits can reject complex documents even when compressed file sizes are small.
• Fields, bookmarks/ranges, notes, comments, tracked changes, content controls, embedded objects, charts, alternate content, picture bullets, legacy VML shapes, external relationships and signed/encrypted/macro packages are rejected. Nothing is fetched or executed.
• Original document properties, thumbnails and unused custom XML are omitted. Retained image/font bytes keep their embedded metadata. This is not redaction or a complete metadata scrub. Existing PNG, JPEG, GIF, BMP, TIFF, WMF and EMF parts are copied unchanged.
• Markup compatibility: moved content must retain the same effective Ignorable namespace behavior. ProcessContent, PreserveElements, PreserveAttributes, MustUnderstand, nonstandard section-property extensions and ambiguous enhanced-style defaults are rejected rather than silently reinterpreted. Inherited xml:space/xml:lang/xml:base on moved document/body/style/numbering containers is unsupported; ordinary explicit run-level xml:space is retained.