From your device. → Via nothing. → To your provider.
Select text in any app, hit a keyboard shortcut, done. Rewrite, translate, grammar-check, summarize, or transform text without leaving what you're doing. Build custom shortcuts with your own prompts, chain multiple steps, and assign each one its own model and account. Text gets processed and replaced in place.
A full floating chat window when you need more than a quick text swap. Attach PDFs, images, Word docs, and Excel files for analysis. Enable web search for real-time information or toggle Claude's extended thinking mode with adjustable reasoning strength. Streaming responses, conversation history, and multi-monitor aware positioning.
A full prompt library with categories, search, and drag-to-reorder. Attach reference materials to prompts for consistent context. Import and export prompts as markdown to share with others. The built-in prompt assistant helps you craft and refine prompts for better results.
Claude (with extended thinking and prompt caching), GPT, Gemini (with thinking levels), Mistral, Perplexity, Ollama for local models, Apple Intelligence on-device, DeepL for dedicated translation, and Stability AI for image generation. Multiple accounts per provider with custom labels and per-account indicator colors.
Personal writing styles shape how each model responds to you. Memory contexts maintain persistent facts and preferences per model. API usage analytics with cost tracking and date-range filtering. Export conversations as markdown, import or export prompts and styles, full data backup and restore. Everything syncs via iCloud.
Bring your own API keys. Your data goes directly to the provider you choose. Nothing passes through our servers.