Skip to main content
StrataNote features an intelligent search system that allows you to instantly find documents and connections between them. The search is optimized for both English and Russian query processing.

​
Opening the Search Panel

The search panel is styled as a Glassmorphism modal window. You can open it in two ways:
  1. Click the Search (Magnifying Glass) icon in the left sidebar.
  2. Press Ctrl + P or Ctrl + K (on macOS: Cmd + P or Cmd + K).

Screenshot: Search Modal Interface

Interface Preview: A blurred modal window overlaying the main app. At the top is the search input field, followed by mode selectors (“Full-Text”, “Semantic”, “By Title”) and a scrollable results list. Key elements: Query input and hotkeys for quickly toggling search modes.

​
Search Modes

You can switch between four search modes depending on your task:

​
1. Full-Text Search (FTS5)

A blazing-fast search across all note contents using SQLite’s FTS5 engine.
  • How it works: Scans the text of notes and outputs direct matches.
  • Features: Generates text snippets (context around the found word) with neon highlights using the HTML <mark> tag.
  • When to use: When looking for a specific word, phrase, command, or exact term in the vault.

Screenshot: Match Highlights in FTS5 Search

Interface Preview: Full-text search results showing text snippets where matching keywords are highlighted with yellow neon backgrounds.
A semantic search that understands the meaning and intent behind your query.
  • How it works: The system converts your query into a vector representation using a local neural model (Xenova/paraphrase-multilingual-MiniLM-L12-v2) and compares it with stored embeddings in the database.
  • Features: Displays a semantic similarity score (e.g., 92% Match). It finds relevant notes even if they do not contain the exact words you typed (e.g., searching for “login” will return notes about “authentication”, “JWT”, or “registration”).
  • When to use: When you remember a concept or topic but forgot the exact words used in the note.

Screenshot: Semantic AI Search Results

Interface Preview: Notes sorted by relevance with purple badges showing the similarity score (e.g., 94% Match).
A classic, instantaneous search filtered strictly by file names.
  • When to use: When you know the exact name of a note and want to open it immediately, bypassing content searches.
Prefix a query with # to match only tags declared inside notes (for example, #project/alpha). Results are limited to notes that actually contain the tag, without matching similar words in the note body. See Tags Explorer & Sidebar for how tags are authored and browsed.

​
Keyboard Navigation

You can fully navigate the search modal without touching the mouse:
  • Arrow Up / Arrow Down — Move selection through the results list.
  • Enter — Open the selected note (restoring tab history and scroll position).
  • Esc — Close the search window.

​
Rebuilding the AI Search Index (For Admins)

While vector embeddings update automatically as you edit documents, bulk imports or external modifications might require rebuilding the index manually:
  1. Navigate to Settings -> Sync tab.
  2. Under the AI Vector Search section, click the “Reindex AI Search” button.
  3. The process runs asynchronously, showing a progress bar (e.g., Indexing files: 45/120... 37%) updated in real-time via WebSockets.

Screenshot: AI Search Indexing Panel

Interface Preview: Synchronization tab in Settings showing an active indexing procedure with a progress bar and loading animations.