Features
Native macOS app (Swift, not Electron)
OCR for scanned PDFs and forms
Multiple AI providers: OpenAI, Azure, OpenRouter, Anthropic, Groq, Mistral, Perplexity, Together AI, Ollama
Local document storage, no cloud uploads
API key stored in Apple Keychain
Workspace switcher for multiple conversations
Text-to-speech and speech-to-text
Export chat to HTML, Markdown, or JSON
Custom system prompt and model parameters
Document viewer with page references
Parallel indexing (configurable up to 5)
Ask questions across uploaded PDFs and documents
Answers with highlighted source citations
Unlimited questions per document
Rewrite summaries until satisfied
Embed PDF AI in any webpage with a single click
256-bit SHA encryption at rest
Role-based access control
OCR for images and scanned text (Team+ plans)
Response personalization (Team+ plans)
Top performer on RepLiQA benchmark
Department & folder level permissions (Team+ plans)
SOC-2 certificate (Enterprise plan)
