- Load Once, Use Anywhere: When you load a model into memory, it stays loaded and ready for immediate use by any feature in the app.
- Centralized State Management: The Model Management Hub is the single source of truth for the state of all your on-device models.
- You’re in Control: From the model management screen, you have granular control over your models.
Getting started
- Introduction
- Getting Started
- The Asynchronous Scribe
- Advanced Transcription Model Management
- The Chat Experience
- Your Conversational AI Companion on iOS
- Execution History & Analytics
- Fine-Tune Your Own AI Models
- The JavaScript Parser Builder
- Jupyter Notebook Integration
- Label Studio Integration
- Local Network RAG
- The Model Management Hub
- Native macOS Experience
- On-Device LLM Integration
- On-Device Translation
- The On-Device Vector Store
- Persistent Toolboxes
- Craft Your Own AI with Personas
- The Private Knowledge Base (RAG)
- Simplified RAG
- Tabbed Navigation
Getting started
The Model Management Hub
Under the Hood
The power and flexibility of the on-device AI in aj is made possible by a core architectural component: the Model Management Hub.
Key Architectural Principles:
Was this page helpful?
⌘I