LogoOctarine Docs
Working with AI

Ask Octarine

Quickly chat with all your notes

Ask Octarine lets you chat with your notes to receive summaries, insights, and answers to your questions. It analyzes your workspace content using a locally-stored embedding model to provide intelligent responses while maintaining complete privacy.

Ask Octarine is only available for users on the Pro License.

Getting Started

Before using Ask Octarine for the first time, the system downloads a small 90MB embedding model to create vector chunks that identify relevant data for your questions. After your workspace is indexed into text chunks, you can start chatting immediately.

  • Access: Click the Ask Octarine sidebar button or press Cmd/Ctrl + O or Cmd/Ctrl + K → Ask Octarine
  • Initial setup:
    • Octarine requires a 90MB model to be downloaded to run embeddings on device (this is fast, cost efficient & privacy friendly since you aren’t using AI credits) for creating smaller chunks of your data.
    • If the setup isn’t downloading correctly, you can press Manual Setup and download the model files from the links and place them in the models directory.
  • Indexing: All workspace notes are processed into searchable text chunks.
    • First time setup will take a few extra seconds since it would index ALL notes in the app into smaller chunks and store it in the local vector database.
    • Post that every new note being created, note being renamed or added to a folder, moved or deleted, would run the embeddings for just those to keep the database in sync.

No data is sent out to any server for embeddings. Only when you use the feature to request a query, your message and any relevant chunks that Octarine determines are sent to your AI provider.

If you wish to keep all things local (even query generation), please use local provider Ollama for your model.

Chat Interface

Each chat session supports multiple messages with different AI models or contexts:

  • Context indicators: Each message displays the context used (if provided) or date range for time-focused queries
  • References: Listed references show which notes were considered for your query
  • Export options: Copy responses as Markdown (for Octarine notes) or plain text
  • Chat titles: Automatically generated after the first response, editable via the Sparkles icon
  • Save the entire chat to a note by clicking on the Create note icon on the chat breadcrumb.
    • User queries are saved a blockquotes
    • Each chat message (user query and AI answer pair) is separated by a line divider.

Modes

  • Ask – Chat with your notes and enjoy clear references and citations.
  • Create – Turn your notes into draft emails, templates, or new notes. You get actionable steps, not just answers.

Narrowing queries: Use the @ symbol to focus on specific folders.

Octarine is also language agnostic, so you can ask in your language of choice, and receive answers back in the same.

Date Range Context

Add explicit date ranges as context for time-specific queries:

  • Click the calendar icon or press / in the input box to open date selection
  • Each date range automatically includes corresponding notes from Daily Desk
  • Selected dates provide temporal context for your questions

Suggested Prompts and History

Empty chats display suggested prompts for quick exploration:

  • Suggested prompts based on random selection of folders with notes
  • Refresh Suggestions button to randomize and display new prompt options

History entries are organized by relative dates:

  • Grouped as Today, Yesterday, 2d ago, or 1w ago instead of absolute dates
  • Stored chats maintain their references when accessed from history
  • Click any history entry to reload the full chat with its context