It uses the recognizer Windows already has
There is no model to download and no image leaves your PC. Windows' own recognizer is called on the spot — it works with the network unplugged.
Find words inside pictures
A scanned contract is a picture, not text. That is why no search tool finds it. CHATDA turns the picture into text and indexes that.
Language packs matter
The recognizer reads the languages installed under Settings > Time & language > Language. If yours is missing, the program says so on screen rather than quietly returning nothing.
It is slow, so it is opt-in
Turning pictures into text takes far longer than reading a document. It is off by default and runs only from the "with image text" shortcut.
From one PC in daily use
Copied from the maker's work PC on 9 September 2026.
- 105,630
- files indexed
- 74,157
- documents read inside
- 102.5 MB
- index size
- 10 ms
- to find 200+ results
Measured on one machine; your numbers depend on your PC and files. The index never leaves your computer.
Frequently asked questions
Does it read handwriting?
Printed text mostly. Handwriting rarely comes out well.
What about skewed scans?
Badly tilted pages get missed. Scanning them straight is the surest fix.