Ghostype is a context-aware AI voice interface for macOS. It knows your active app to automate the workflow: ⚡️ Smart Auto-Send: Detects context to format & execute "Send" (e.g., Cmd+Enter). 🎨 App Profiles: Auto-switches tones per app. 🧩 Ghost Morph: Custom Agent Skills. 👻 Ghost Twin: Learns your style locally.
Hey Product Hunt! 👋
I’m the maker behind Ghostype.
I built this because I got tired of the "AI Accent." I use voice dictation daily, but I spent more time editing the output to sound like me than I did actually speaking. Plus, having to manually hit Cmd+Enter after every dictation felt like a broken workflow.
So I built Ghostype—a macOS-native layer that sits between your voice and your apps.
It’s not just "speech-to-text." It’s an automation engine:
⚡️ Smart Auto-Send: My favorite feature. It detects if you're in Slack, Discord, or Notes, and executes the correct "Send" command automatically. No keyboard needed.
🎨 App-Specific Profiles: It knows context. It keeps my tone "Professional" in Outlook but switches to "Casual & Lowercase" in Telegram.
🧩 Ghost Morph: I added custom Agent Skills. Hold a modifier key to turn a rambling voice note into a structured Jira ticket or a Twitter thread instantly.
👻 Ghost Twin: The "magic" part. It analyzes your local writing history to learn your specific style/slang, so the output feels like you, not a robot.
🔐 A note on Privacy: I’m an indie dev and literally cannot afford the server costs to store your data. 😂 So, Ghostype is built with a Local-Storage architecture. Your history and style vectors stay 100% on your Mac.
I’m currently in Beta and iterating fast. I’d love to hear your feedback on the "Smart Send" logic—does it fit your workflow?
Thanks for checking it out! 👻
Dawei Geng