Why extracting beats right-click and save
Office files are ZIP containers, and every inserted picture is stored inside as a complete, untouched
image file โ in word/media/, xl/media/ or ppt/media/. Extraction
copies those bytes out verbatim.
Saving a picture from inside Word does something different. Word hands over the image as currently displayed, which frequently means the compressed, downscaled version Office created when the document was saved with picture compression enabled. A 4000-pixel photograph inserted into a document can come back as 1200 pixels that way. The copy in the media folder is usually the original.
For a full technical breakdown of every part inside a .docx archive, including how images are stored, read how DOCX files store content.
How to extract images from a Word document
- Open the document. Drop the .docx, .xlsx or .pptx onto the box above, or click to browse. Only the media folder is read, so a large document opens quickly.
- Review what is inside. Every image is listed largest first, with its pixel dimensions, format and file size, so the full-resolution originals are easy to spot.
- Save what you need. Click Save on any image. The bytes written are the bytes stored in the document โ no conversion step happens.
What comes out
- Every stored image, in original format and at original resolution.
- Pixel dimensions, read straight from each file's own header โ PNG, JPEG, GIF, BMP and WebP โ without decoding the whole picture.
- Format and byte size for each, sorted largest first, because the largest files are almost always the ones worth having.
- A preview of anything the browser can display.
The same tool works on all three Office formats, because they use the same container layout. Point it at a spreadsheet or a deck and it reads their media folders instead.
Which files and image types work
Accepted documents are .docx, .xlsx, .pptx and their
macro-enabled variants .docm, .xlsm and .pptm. Files exported by
Google Docs, LibreOffice and Keynote work too.
PNG, JPEG, GIF, BMP, WebP and SVG images preview and save normally. EMF and WMF metafiles โ the vector format Office uses for pasted charts and shapes โ cannot be displayed by a browser, so they are listed without a preview and flagged; they still save correctly and open in Office, Inkscape or LibreOffice.
Why the document is never uploaded
The archive is read from your disk by the browser and decompressed with the browser's own DecompressionStream. No server takes part, so the document is never transmitted and no copy exists after you close the tab.
This matters more for images than for text. Photographs embedded in documents frequently carry EXIF metadata including GPS coordinates and camera serial numbers, and an image uploaded to an extraction service takes all of that with it. Extracted locally, the file goes straight to your disk.
What is not extracted
Only genuinely stored pictures come out. Four things do not:
- Shapes, arrows and text boxes. These are drawing instructions, not images, so there is nothing to save.
- Charts. A native Office chart is data plus a rendering rule. Only charts pasted as pictures exist as files.
- SmartArt. Generated at display time from a layout definition.
- Cropping and effects. Cropping in Word hides part of an image without altering the stored file, so an extracted picture may show more than the document does. That is usually an advantage, occasionally a surprise.
Images are also not renamed to anything meaningful, because the document does not record a meaningful
name โ image1.png is genuinely what is inside.
Who extracts images from Office files
- Designers โ recovering original artwork from a document when the source files are lost.
- Content teams โ pulling every photograph out of a supplied document for a CMS upload.
- Anyone rebuilding a document โ extracting assets to reuse in a new template.
- Auditors and investigators โ recovering the images in a file at full resolution, including anything cropped out of view.
- Report writers โ reusing charts and diagrams that were pasted in as pictures.
This tool compared with unzipping the document yourself
Because a .docx is a ZIP, you can rename it to .zip and open it by hand โ the images are
in word/media/. That works, and it is worth knowing. This tool does the same thing without
renaming anything, and adds what a file manager will not tell you: pixel dimensions, previews, and a
largest-first ordering that puts the originals at the top.
To see every part inside the container rather than just images, use the ZIP extractor. For the document's words, use the DOCX text extractor. To read the camera and GPS data an extracted photograph carries, follow up with the OCR tool for text in the picture.
For a step-by-step walkthrough, read how to extract images from Word. If your file is hosted in Google Drive, see how to download images from a Google Doc.
DOCX image storage and extraction edge cases
Office images are stored as relationship targets. An
<a:blip r:embed="rIdN"/> element in drawing markup resolves through the
part's .rels file to a path like word/media/image1.png. The file
extension in the media part determines the format reported.
Three edge cases: (a) EMF and WMF files (Windows Enhanced Metafile and Windows Metafile)
are vector formats stored as opaque binary blobs โ they are saved as-is and cannot be opened
in a browser without a dedicated viewer; (b) the thumbnail at
docProps/thumbnail.jpeg is a cover-page preview image and appears as a separate
"thumbnail" entry, distinct from images embedded in the document body; (c) images in headers,
footers and footnotes live in their own XML parts with their own relationship files โ the
extractor walks all relationship targets across all parts, so those images are included even
though they sit outside the main document body.
Frequently asked questions
How do I extract all images from a Word document at once?
Drop the .docx onto this page. Every image is listed with a Save button, largest first. Browsers restrict saving many files in one action, so images are saved individually.
Will the images lose quality?
No. The bytes saved are the bytes stored in the document. There is no re-encoding step, which is why this often gives a higher-resolution file than saving the picture from inside Word.
Is my document uploaded to a server?
No. The file is read from your disk and unzipped by your browser. It is never transmitted.
Does it work on Excel and PowerPoint too?
Yes. All three formats store images the same way, so a .xlsx or .pptx works identically.
Why can't I preview some images?
They are EMF or WMF metafiles, a Windows vector format browsers cannot display. They are listed and flagged, and they save correctly for use in Office, Inkscape or LibreOffice.
Why does an extracted image show more than the document?
Cropping in Office hides part of a picture without changing the stored file. The extracted original still contains the cropped-out area.
Why are the filenames just image1, image2?
That is what the document stores. Office renames pictures on insert and does not keep the original filename anywhere in the file.
Can I get images out of an old .doc file?
Not with this tool. The legacy .doc format is binary and does not store images as separate files. Open it and save as .docx first.