Your Personal Voice Interface for Any App

An open-source platform for AI voice dictation that you can customize and personalize to your exact needs. Choose your models, write your prompts, make it yours.

Get Started Star on GitHub
— stars
Windows macOS Works with any text input
Demo of Tambourine open source voice dictation: speaking into microphone and text appearing in Windows Notepad

What People Say

"I love this kind of project: open source, hackable, filling a gap in the voice ecosystem with a really useful codebase, pushing the boundaries of what you can do with 'voice AI engineering.'"

Kwindla Hultman Kramer Co-founder / CEO, Daily (Pipecat)

"When I saw this I knew we had to give it a try and it works brilliantly."

Stuart Wood Speechmatics

"Most of us speak around 130-160 wpm and type 40-50 wpm. That means speaking is about 3x faster than typing. I've been using Tambourine to dictate directly to Claude Code. I'm impressed."

Derek Haynes Finchvox

The Foundation of Voice Interfaces

You can build on top of Tambourine to create exactly the voice experience you need.

Choose Your Models

Pick the STT and LLM that work for you. Use cloud providers or run fully local with Whisper and Ollama.

Write Your Prompts

Customize how your speech gets formatted. Add your own rules, adjust the tone, define how punctuation works.

Build Your Features

Implement features like backtracking corrections and list formatting your way. The logic is in prompts you control.

Personalize Everything

These features ship by default, but they're implemented in customizable prompts. Modify them, extend them, or build your own.

Personal Dictionary

Teach it technical terms, names, and jargon specific to your work.

"ant row pic" → Anthropic

Backtrack Corrections

Say "actually" or "scratch that" to correct mistakes as you speak.

"at 2 actually 3" → at 3

Smart List Formatting

Automatically format lists.

"one eggs two milk" → 1. Eggs 2. Milk

Your Own Ideas

The prompts and logic are yours to modify. Build what you need.

Edit prompts in Settings → Prompts

Coming Soon: Context-Awareness

Tambourine will access surrounding context to make dictation smarter.

App-Dependent Formatting

Automatically adapt to the app you're in. Emails get salutations, code editors get syntax-aware output.

Selection Commands

Highlight text and say "make this more formal" or "translate to Spanish" to transform it in place.

Voice Shortcuts

Create custom triggers like "insert meeting link" that expand to frequently used text.

Get Started

Run it yourself for free, or let us handle the infrastructure.

Available Now

Self-Host

Run the open source code yourself. Full control, completely free, requires some setup.

View on GitHub
Coming Soon

Hosted Service

We run the infrastructure for you. No setup, no fuss, just download and start dictating.

Join Waitlist
AGPL-3.0 License

Built in the Open

Tambourine is fully open source. Contribute features, report bugs, or join the community. The platform is yours.

Star the Repo Join Discord Contribute