dictare 0.2.2

Voice-first control for AI coding agents

DICTAre™

Voice layer for AI coding agents.

Speak to your agent. No window focus required. 100% local.

dictare.io · OpenVIP™ Protocol

If you want to know how a poker game turned into a voice interaction system for coding agents... watch this →

---

Why dictare

Most voice tools (Wispr Flow, Superwhisper, etc.) simulate keystrokes — they type into whatever window has focus. Switch to your browser and your code gets your voice.

Dictare uses a protocol. Your agent listens via SSE and receives transcriptions regardless of window focus. Your coding agent can be behind 3 other windows — it still gets your words.

Features

• No focus required — agent receives voice even when its window is in the background
• Agent-native — transcriptions go to the agent protocol, not a text field
• 100% local — STT runs on-device, zero data leaves your machine
• Multi-agent — switch agents by voice: "agent coding", "agent review"
• Open protocol — OpenVIP — any tool can implement the SSE endpoint
• Bidirectional — STT (voice in) + TTS (voice out)

Install

macOS — full guide

brew install dragfly/tap/dictare

Linux — full guide

curl -fsSL https://raw.githubusercontent.com/dragfly/dictare/main/install.sh | bash
sudo usermod -aG input $USER   # required for hotkey (log out/in after)

Permissions

macOS — grant when prompted:

1. Microphone — prompted on first launch
2. Input Monitoring — System Settings → Privacy & Security → enable Dictare
3. Accessibility — needed for keyboard mode (typing into other apps)

After granting all three: dictare service restart

Linux — two steps:

1. Input group (hotkey, X11 + Wayland): sudo usermod -aG input $USER — log out/in
2. ydotool (keyboard mode on Wayland): sudo apt install ydotool

Quick Start

dictare agent freddie       # starts the default profile (Claude Code)

That's it. The service starts automatically. Speak — your agent receives the transcription.

If you prefer a different coding agent:

dictare agent ozzy --profile codex      # OpenAI Codex
dictare agent gilmour --profile gemini  # Google Gemini CLI
dictare agent bowie --profile aider     # Aider

How It Works

  Microphone
      │
      ▼
  STT Module       Whisper (MLX / CTranslate2) or Parakeet (ONNX)
      │             all local, zero cold-start
      ▼
  Pipeline         submit detection, mute control, agent switching
      │
      ▼
  OpenVIP          HTTP / SSE — open protocol
      │
      ▼
  Agent            receives transcription, no window focus needed

The engine runs as a background service (launchd on macOS, systemd on Linux). STT models are preloaded at startup. Each agent connects in its own terminal.

Agent Profiles

Profiles are predefined in ~/.config/dictare/config.toml:

[agent_profiles]
default = "claude"

[agent_profiles.claude]
command = ["claude"]
description = "Claude Code"

[agent_profiles.codex]
command = ["codex"]
description = "OpenAI Codex"

[agent_profiles.pi]
command = ["pi", "--provider", "ollama", "--model", "qwen3:8b"]
continue_args = ["-c"]
description = "Pi + Ollama local, agentic with tools"

Then connect:

dictare agent freddie                      # default profile (claude)
dictare agent ozzy --profile codex         # use codex profile
dictare agent -- claude --model opus       # explicit command override

Voice Commands

Say	Action
"ok, submit" / "ok, send" / "ok, invia" / "ja, senden"	Submit to agent (Enter)
"ok, mute" / "ok, hold on"	Mute (stop listening)
"ok, listen" / "ok, listen up"	Unmute (resume listening)
"agent coding" / "agent review"	Switch active agent

Submit triggers are multilingual (en, de, es, it, fr) and fully configurable.

Hotkey Cheat Sheet

Default hotkey: Right ⌘ (macOS) / Scroll Lock (Linux).

Gesture	Action
Single tap	Toggle listening on/off
Double tap	Submit (send Enter to agent)
Right Alt + hotkey	Switch mode: agents ↔ keyboard

Service Management

dictare service install     # Install + enable (auto-starts at login)
dictare service start       # Start the service
dictare service stop        # Stop the service
dictare service restart     # Restart the service
dictare service status      # Show service and engine status
dictare service logs        # View recent logs
dictare service uninstall   # Remove the service

Keyboard Mode

No agent? Use dictare as a dictation tool — voice to keystrokes in any app.

dictare config set output.mode keyboard

Hotkey to toggle listening (configurable):

• macOS: Right ⌘ by default
• Linux: Scroll Lock by default

dictare config set hotkey.key KEY_RIGHTALT   # change hotkey

Text-to-Speech

dictare speak "Hello world"
dictare speak --engine piper "Hello"
echo "Hello" | dictare speak

Engines: espeak, say (macOS), piper, kokoro

Configuration

dictare config edit           # Open config in editor
dictare config list           # Show all settings
dictare config get stt.model
dictare config set stt.language it

Full configuration reference at dictare.io/docs/configuration.

Development

git clone https://github.com/dragfly/dictare && cd dictare

# macOS Apple Silicon (MLX GPU acceleration)
uv sync --python 3.11 --extra mlx

# macOS Intel / Linux
uv sync --python 3.11

# Run engine in foreground
uv run --python 3.11 dictare serve

# Tests
uv run --python 3.11 pytest tests/ -x

# Tests (parallel)
uv run --python 3.11 pytest tests/ -x -n auto

Ghostty users: add keybind = shift+enter=text:\n to config. See TERMINAL_COMPATIBILITY.md.

Protocol

dictare is the reference implementation of OpenVIP — an open protocol for voice input to AI agents. Any tool can implement the SSE endpoint and receive voice transcriptions from dictare.

License

MIT