Microsoft Edge WebView2 Caches
Many Windows apps embed Microsoft Edge WebView2 to render sign-in screens, settings panes, help pages, and full web interfaces using the Edge/Chromium engine. Those embedded views generate HTTP disk cache entries, compiled V8 code cache, GPUCache data, shader blobs, and custom dictionary cache files to speed up page loads, script startup, graphics rendering, and spellchecking. Kudu removes only those WebView2 cache directories under Local AppData, leaving cookies, sessions, IndexedDB, Local Storage, service-worker storage, offline content, accounts, and app settings untouched.
Why clean Microsoft Edge WebView2 Caches?
- Stale HTTP cache entries can keep an embedded app showing an old login page, missing styles, or broken buttons even after the site was updated
- Invalidated V8 code cache after a WebView2 runtime or app update often causes unusually slow first opening of embedded pages until the bytecode cache rebuilds
- GPUCache data from a previous graphics driver version can become invalid, leading to black or blank WebView2 panes, flicker, or rendering glitches until the GPU cache is cleared
- Corrupted shader blobs can cause stutter when animated dashboards, charts, or video-heavy web views first appear, then improve after the cache is regenerated
- An oversized disk cache under Local AppData can consume hundreds of megabytes across multiple apps that host WebView2, with no effect on saved sign-ins or local app data when cleaned
- Damaged dictionary cache files can make spellcheck suggestions disappear or behave inconsistently inside text fields rendered by WebView2
- Repeatedly reused stale web assets in the cache can leave embedded help, account, or settings pages loading incorrectly while the rest of the app appears normal
Cache paths Kudu targets
Windows
%LocalAppData% |
Common questions about Microsoft Edge WebView2 Caches
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.