Skip to content

FAQ

Pia (Personal Intelligent Assistant) is a Windows desktop AI assistant that supports multiple AI providers, local speech processing, cloud sync with end-to-end encryption, and self-hosting.

Pia has a free Community tier for self-hosted usage and a Free cloud tier with basic features. Pro and Enterprise plans offer additional features like E2E encryption and team management.

Currently Pia is available for Windows 10+ (x64). macOS and Linux support is planned for the future.

OpenAI, Azure OpenAI, Anthropic, Mistral, and Pia Cloud (server-proxied). See the AI Providers Guide.

Yes. When using a direct API key, conversations go straight from your device to the AI provider. With Pia Cloud + Pro plan, conversations are end-to-end encrypted before leaving your device.

Only if you opt into Cloud Sync. Local-only mode sends nothing to Pia servers. Your API keys are always stored locally.

Yes. See the Self-Hosting Reference for Docker Compose setup instructions.

Pia uses Velopack for auto-updates. It checks for new versions on startup and applies delta updates automatically.

Pia stores data in %APPDATA%/Pia/ including your local database, settings, and Whisper models.

Yes. Pia has a built-in Kanban board with drag-and-drop reordering, columns, priority levels, and voice input. See the Todos & Kanban Guide.

Memory lets you store structured personal knowledge — contacts, projects, notes — that the AI assistant can reference during conversations for more relevant responses. See the Memory Guide.

In the Assistant chat, type @ followed by a domain like @Memory, @Todo, or @Reminder to reference specific items directly in your message. An autocomplete popup helps you find the right item. See the Memory Guide for details.