Apps
fontconfig
Fontconfig keeps a generated font discovery cache under %LocalAppData%/fontconfig/cache so applications can avoid rescanning every installed font file on each launch. Those cache files store parsed font metadata such as family names, style and weight information, character coverage, file paths, and timestamps used during font matching and fallback. Kudu removes only these regenerated fontconfig cache files, not your installed fonts, app settings, accounts, or documents.
Why it matters
Why clean fontconfig?
- Stale font directory metadata after installing or removing fonts can make apps show missing families, duplicate entries, or old font names until fontconfig rescans
- Cached character coverage and fallback data that no longer matches the current font set can cause tofu boxes or the wrong fallback font for CJK, emoji, or symbol text
- Moved or deleted font files leave behind cached paths, which can produce slow startup pauses while apps repeatedly probe missing files before rebuilding the cache
- Corrupted cache records can break font matching and show the wrong weight or style, such as bold text rendering as regular or italics not being picked correctly
- After replacing a font with a newer build, outdated cached properties can make applications keep using the previous metadata, so the new family or variable axes do not appear immediately
- Large, long-lived caches can accumulate obsolete entries from old font scans, increasing disk use and making the first font lookup in compatible apps feel slower than it should
- SQLite-style database cleanup does not apply here because fontconfig uses generated binary cache files rather than a relational store, so deleting them forces a clean rebuild instead of vacuuming fragmented pages
What gets cleaned
Cache paths Kudu targets
Windows
%LocalAppData%/fontconfig/cache |
Frequently asked
Common questions about fontconfig
Related cleaners
Free & open source
Download Kudu and reclaim your disk space.
Available on Windows, macOS, and Linux. No account required, no feature gates, no telemetry without consent. All cleaning targets are open source and community-auditable.