Search
Search covers text queries over filenames, titles, and descriptions, plus a People filter for narrowing results to a recognized person.
Structured filter tokens can be typed directly into the query box: person:"Name" (People filter), year:2023 and year:<2020/year:>2020 comparisons, from:/to: date ranges, album:"Name", filename:"name.jpg" (exact, case-insensitive filename match, extension included), favorite:true|false, approval:tentative|approved, face:detected|face:unassigned and date:none. Token names are case-insensitive; quoted values allow spaces. Filters AND-combine (filename:"IMG_0001.jpg" person:Alice finds that one photo if it contains Alice); a token matching nothing reports an unmatched chip and returns no results.
When compatible local embedding models are available, Memories also ranks results with local semantic search. The supported pairing is a ViT-B-32 image and text embedding model. No media or query content is sent to a cloud service.
- Semantic search reports itself unavailable when stored image embeddings and query embeddings come from different model name or version provenance, rather than returning misleading results.
- The People filter links to People and Faces.
- Results open in The Media Viewer, preserving the query.
Everything runs against the local library metadata and local embeddings, so search continues to work offline inside your own deployment.
Details
Memories Wiki
Using Memories
- Accounts, the Library Hub, and Data Export
- Albums and Automatic Suggestions
- Explore and Discovery Domains
- The Media Viewer
- People and Faces
- Places and Map
- Plants (In Development)
- Datenschutzerklärung (Privacy Policy)
- Fast, Progressive Image Loading
- Search
- Sharing and Households
- Timeline and Photos
- Uploading and Import
Self-Hosting
- Backup and Restore
- CLI Password Reset
- Deployment Architecture
- Known Issues and Operator Diagnostics
- Map Tiles and Place Data Privacy
- Storage Layout
- Upgrades and Versioning
- Workers
Reference