Index status and failed documents

Check progress

Open Indexing in the sidebar.

• Overall progress and counts for found · succeeded · skipped · failed
• Searchable now: documents already read are searchable even before indexing finishes.
• Embedding (Standard and up): the step that powers search by meaning. It shows the documents left and the time left at the current speed (for example "27,842 left · about 3 h 41 min"). With tens of thousands of documents this normally takes hours; keyword search works in the meantime. If it is resting because the PC is busy, the reason is shown (for example "paused — running on battery"). Source code and config files (.java, .js, .json, etc.) skip this step and are found by keyword only.
• "Pause", "Resume" and "Stop" buttons — stopping pauses reading, embedding and AI understanding together.

The top bar always shows indexing status and the document count.

Status of each document

Folders shows a status for each document.

Documents with no status are fully indexed.

Fix failed documents

1. Open Indexing › "Failed documents".
2. Check the files in the list. "Copy path" copies a file's location.
3. Fix the cause, then select "Retry failed documents". A failed file isn't retried on every check by itself — retrying reads it again.

Common causes

• Files that ask for a password to open, or damaged files (PDFs that only restrict editing are read)
• Another program has the file locked
• The file is larger than the maximum file size → raise it in Settings › Indexing and re-index
• Scanned PDFs → reading their content needs OCR on Pro

A new file doesn't show up in search

1. Check in Indexing whether indexing is still running.
2. Make sure the file isn't in an excluded folder or extension.
3. In Folders, run "Re-index changes" on that folder.

Still stuck? See Troubleshooting.