Intercepts file uploads on Claude and ChatGPT and converts them to Markdown via the MarkItDown backend, to save tokens.
Users
53
+13 in 30 days
Rating
5.0 / 5
unchanged in 30 days
Ratings
1
0 in 30 days
Store rank
#107K
#89K among extensions
Version
0.1.2
updated 10 Aug 2026
Users
monthly, 3 months
Latest, 18 Oct 2026:53 usersMove the mouse over the chart for any month
Ratings count
monthly, 3 months
Latest, 18 Oct 2026:1 ratingsMove the mouse over the chart for any month
The Chrome Web Store rounds user counts down to the leading digit, so the user line moves in steps. The ratings count is exact and is the better signal of real adoption.
About Tokens Saver
Note: By design, The extension only works with the drap and drop.
If you want to upload a document without the extension, use the "+" icon instead of the drap and drop feature
Tokens Saver converts your documents to Markdown BEFORE they reach the AI chat so you send lean, clean text instead of a heavy file, and your token usage (and usage limits) stretch much further.
HOW IT WORKS
Drag & drop a document onto claude.ai or chatgpt.com. Tokens Saver intercepts it, converts it to Markdown using Microsoft's MarkItDown converter, and inserts the result straight into your prompt box. The raw file never gets uploaded to the chat site.
YOU CHOOSE, PER UPLOAD
• Drag & drop (or paste) a file = converted to Markdown, tokens saved
• The site's own + attach button = normal upload, extension stays out of the way
• Images are never touched — the AI's vision should see those as-is
SUPPORTED FILES
PDF, DOCX, PPTX, XLSX, CSV (up to ~4 MB)
WHY MARKDOWN?
AI chat services process raw files page-by-page, sometimes even as images which burns tokens fast. Plain Markdown text carries the same content at a fraction of the cost, and models read it extremely well.
PRIVACY
Files you convert are processed in memory and never stored. No analytics, no tracking, no accounts. Open source (MIT): https://github.com/Zaki-Haider1/Tokens-Saver
NOTES
• You can also self-host your own backend. (See GitHub Page)
• Scanned/image-only PDFs are not supported yet (no OCR).
• Early release feedback and contributions welcome on GitHub.
DISCLAIMER:
Try to use the extension mainly for files which mostly have text, since the extension does not support and hence ignores direct images. If you file is too big or there are any other problems, use the "+" icon to upload your documents instead of the drag and drop feature.