Mic to Clipboard

Live

Speak once, paste anywhere

Built by · Launched April 2026

Download on App Store
Mic to Clipboard screenshot
ios react-native expo ios speech-recognition typescript

What is it?

Mic to Clipboard is a dead simple iOS app: tap the mic button, say what you want to type, and the transcription lands in your clipboard. Paste it anywhere. Claude Code, Slack, a text message, a search box. Whatever you were about to type out.

One screen. One button. Nothing to configure.

Why it exists

Claude Code on Bedrock doesn’t expose microphone access. The native Claude desktop client does, but through work I use the Bedrock version, which meant typing every prompt by hand. Over time that friction added up. This app fills the gap.

It also works for anything else where speaking is faster than typing. Long Claude prompts, emails you’d rather dictate, notes to yourself while your hands are full.

How it works

Tap the mic button. Speak. The app uses Apple’s on-device speech recognizer (SFSpeechRecognizer) so nothing leaves your phone. When you stop, the full transcript is written to the clipboard and the button turns into a checkmark. Switch to wherever you want the text and paste.

The transcription handles natural pauses in speech, accumulating partial results into one complete string before copying. Interim words appear on screen as you speak so you can see it working in real time.

Features

  • On-device transcription using Apple’s speech recognizer, no data sent to external servers
  • One-tap workflow with no settings to configure before use
  • Real-time display shows words appearing as you speak
  • Haptic feedback on copy confirmation
  • Light and dark mode support
  • Available on iPhone, iPad, and Apple Silicon Macs via the Designed for iPad compatibility layer

Tech

React Native with Expo, using expo-speech-recognition to wrap Apple’s SFSpeechRecognizer and expo-clipboard to write the result. The app checks at runtime whether the device supports on-device recognition and falls back to network-based transcription on older hardware. Built and shipped with EAS.

Read the full build log for the technical details, including how continuous speech recognition accumulates across pauses and how App Store submission actually went.