Speech to text.
Capture dictation and meeting speech across supported languages and accents, including the moments they mix.
“Order ready hai?One conversation, two languages
I’ll pick it up at six.”
Understand speech across accents and languages.
Get clear transcripts and natural translations.

Hear what happens.
People switch languages mid-thought.
Pick a sample. Press play. See it understood.
Two languages. One thought.
“Meeting kal 3 baje hai. Please send the notes.”
Original languages preservedA hands-on example of the experience. This demo does not call the live API.
Local notes.
Global conversations.
translation targetsSource-language coverage varies by mode
A place for the whole conversation.
From a quick thought to a multilingual meeting, keep the original words, their meaning, and your own notes together.
Capture dictation and meeting speech across supported languages and accents, including the moments they mix.
“Order ready hai?One conversation, two languages
I’ll pick it up at six.”
Follow the meaning in another language, with translations kept beside the original speech.
“Necesito cambiar la entrega.”
“I need to change the delivery.”
Keep your own notes, then create a separate summary with the decisions and next steps that matter.
The words. The meaning. The next step.
Write in your own voice. Keep the source speech close. Let a separate summary help you find what matters.
Meet the notebooktexts, side by side
summary formats
bots joining your call
Original speech and translation stay together. Choose meeting, sales, or support summaries. Recording starts when you choose.
A little setup. A lot less typing.
Start with a note. The first-run guide helps you add speech, Apple Calendar, and call suggestions when you want them.
Apple Silicon, macOS 26+. Choose DMG or Homebrew.
Follow the short guide, or skip straight to notes.
Write first. Add speech and Calendar whenever you’re ready.
Apple Silicon · macOS 26+ · 0.2.0-preview.1
Unnotarized preview. macOS may block first launch. Review Apple’s installation guidance and the release notes and limits.
SHA-256 checksumsbrew install --cask rirachii/tap/chirpberryOpen your notebook. Make room for a thought.Install into Applications → Chirpberry Electron Candidate. Preserve an existing app and never replace it while running. Speech needs your own Valsea account and credits; Ask meeting needs a separate OpenAI key.
Requires Xcode 26+ with command-line tools, Git, and Node.js 22.12+.
git clone https://github.com/rirachii/chirpberry.git chirpberry
cd chirpberry
git checkout --detach 040e04e9f69ee96f6c06db70320ff1aa9a988d0f
npm ci --prefix desktop
npm start --prefix desktopMake room for voice.
Start with your voice.
Keep the words you need.
Turn a passing thought into text. Dictate, copy the final words, and keep a Scratchpad note for later.
Follow conversations across supported languages, with the original transcript and its translation together.
Keep personal notes, create a separate summary, and take the next steps with you.
Simple, from the start.
The app is free software.
Cloud speech uses your own account.
Write, organize, search, and export your notes locally.
Connect your account for transcription, translation, and enhanced notes.
Chirpberry is powered by Hushfig. The current source build connects to the licensed Valsea service using your own account and credits. Microphone plus computer audio uses two streams; optional speaker detection adds provider cost.
A notebook that stays yours.
Your saved notes and transcripts stay on your device. During capture, speech goes to Valsea for processing. Chirpberry does not save audio recordings.
Read how it works