rar and 7z archives, manual tagging, and duplicate fixes
Meshory v1.6.0 · Aug 14, 2026 · 26 changes.rar and .7z archives.
Meshory previously only read inside .zip. RAR and 7z archives are now first-class in the same way: indexed, their contents promoted into the library as real models, with thumbnails, search, filters, facets, stats, duplicate detection and custom thumbnails all working on the files inside them. Extraction runs through a bounded cache, 5 GB by default, that you can resize or clear in settings.
Live duplicate analysis.
The Duplicates pass now shows its groups as it finds them instead of only at the end, reports progress in the sidebar, and can be stopped without losing what it already found. Duplicate detection stays on by default and can now be switched off in settings, where you can also clear saved results.
Manual tagging.
Add your own tags to models from the right-click menu, the details panel, or across a bulk selection, with recently used tags suggested first.
Drag out of library.
Drag one or more selected models straight out of Meshory into another folder or app, with a thumbnail following your cursor.
Archive preview thumbnails.
Models packed inside archives now get thumbnails, including OBJ and FBX files, loading progressively as Meshory reads through each archive rather than all at the end.
Folder exclusion.
Exclude specific folders inside a library source from being scanned, right from the sidebar. Adding a source that overlaps another is now blocked instead of silently importing an empty folder.
Trash and remove sources.
Trash models straight from the right-click menu, with a warning before trashing an archive, and see progress or any errors while a source is being removed.
Database recovery.
If your library database gets damaged, a guided recovery window walks you through restoring it. Downgrading to an older version now restores a compatible database automatically.
Diagnostics privacy.
Diagnostic reports now hide your file and folder names by default and include the scanner's current state to help with troubleshooting.
Per-source tag pausing.
Pause automatic tag and metadata generation for an individual source, with progress shown while it runs.
Cloud file scanning.
Library scans now pick up Windows cloud placeholder files, such as those created by OneDrive, instead of skipping them.
Faster first scans.
Scanning a library for the first time completes faster.
Faster tagging.
Automatic tag generation runs faster.
On-demand previews.
Previews are now generated only when you need them instead of upfront for every model, reducing background work.
Geometry duplicates were being missed.
Any model with a file size unique in your library was dropped from geometry grouping entirely, even though it had a valid geometry fingerprint. The same mesh exported from two different tools is exactly the case most likely to have a unique size, so the geometry tier was under-reporting the matches it exists to find. Re-run the Duplicates pass to pick up what earlier runs missed.
Trashing a geometry duplicate group did nothing.
Move to Trash on a geometry-matched group was silently refused, because the safety check that keeps one surviving copy compared files by content rather than by geometry.
Stop button behavior.
Stopping a source only paused scanning before, leaving preview generation running in the background.
Stuck scanning indicator.
After certain refreshes, the app could stay stuck showing a scan in progress that had already finished.
Missing thumbnails after scans.
Thumbnails sometimes stopped generating after a scan finished until the app was restarted.
Incomplete model details.
Opening a model before it had a preview could leave its details permanently incomplete.
Detail page refresh.
The open detail page didn't update to show a preview once it finished generating.
Stale archive contents.
Re-saving an archive in place could leave its old contents showing forever instead of the updated ones.
High zoom text clipping.
Detail values were clipped instead of wrapping at high text zoom levels.
Folder sort order.
Folder listings sorted names case-sensitively, putting capitalized names out of order.
Search after deletion.
Deleted models could still show up in search results.
Flat view filters.
Filtering by file format in the flat view could miss files it should have shown.