Repository navigation
Releases
Voice Command AI 0.1.1
Voice Command AI 0.1.1 adds the Linux desktop app (x86_64 and ARM64). Windows and macOS get the same version with the fixes below; installed 0.1.0 apps offer the update in Settings → About.
Downloads
| Platform | File |
|---|---|
| Windows 10/11 (x64) | Voice.Command.AI_0.1.1_x64-setup.exe, signed by MetaVu Co., Ltd. |
| macOS, Apple silicon | Voice.Command.AI_0.1.1_aarch64.dmg, Developer ID signed and notarized |
| macOS, Intel | Voice.Command.AI_0.1.1_x64.dmg, Developer ID signed and notarized |
| Linux x86_64 | Voice.Command.AI_0.1.1_amd64.deb, Voice.Command.AI-0.1.1-1.x86_64.rpm |
| Linux ARM64 | Voice.Command.AI_0.1.1_arm64.deb, Voice.Command.AI-0.1.1-1.aarch64.rpm |
Every desktop app includes the VCA server (vcad) and the vca command. Phone apps: Google Play; the iPhone/iPad app is in App Store review.
Linux
.deb(Debian, Ubuntu):sudo apt install ./Voice*.deb..rpm(Fedora):sudo dnf install ./Voice*.rpm.- Needs Ubuntu 22.04, Debian 12, Fedora 36 or newer (glibc 2.35+, WebKitGTK 4.1).
- The in-app updater installs newer packages of the same type and asks for your password through
pkexec. - On Wayland, restored terminal windows open but are not moved; arranging windows needs X11 with
wmctrlorxdotool. See the user guide.
Fixed
- macOS: pairing a phone from the
vcacommand could not finish; the server now identifies the local caller correctly. - Closing the chat search right after starting Search earlier messages now always stops loading earlier pages.
- Linux: the fallback key folder (when no Secret Service runs) is readable only by your user.
- Linux: session discovery finds agents whose executable name is longer than 15 characters.
Full list: CHANGELOG.
Voice Command AI 0.1.0
The first public release of Voice Command AI: one voice to command every AI coding agent.
Voice Command AI brings the AI coding agents running on your own computer (Claude Code, Codex CLI, Gemini CLI, Antigravity and more) into one command center. Speak or type commands, approve decisions from your phone or tablet, get alerts and spoken briefings, and talk to the Command Center, an AI conductor that briefs you on every session.
Download
| Platform | File |
|---|---|
| Windows 10/11 (x64) | Voice.Command.AI_0.1.0_x64-setup.exe (code-signed) |
| macOS, Apple silicon | Voice.Command.AI_0.1.0_aarch64.dmg (notarized) |
| macOS, Intel | Voice.Command.AI_0.1.0_x64.dmg (notarized) |
| iPhone and iPad | App Store (in review) |
| Android | Google Play (in review) |
Linux packages come in a later version.
Highlights
- Every agent session on your computer as a live card, with alerts for decisions and questions.
- Voice dictation, hands-free mode and the "Voco" wake word; spoken alerts and briefings.
- Pair a phone, tablet or another computer by QR code. End-to-end encrypted (Noise), no account, no VCA cloud. Connect from anywhere over Tailscale, Cloudflare Tunnel, DNSPilot or your own provider.
- Command Center: an optional AI conductor that briefs you on all sessions and drafts messages for your approval.
- Save and restore windows and screen layouts; to-dos from agents; resource admission so parallel builds don't stall your machine.
- Security: authenticator (TOTP) pairing, per-device permissions, audit log and a kill switch.
- Try it without a computer: the mobile apps have a built-in demo.
Walkthrough video: https://youtu.be/y1nxSgDrQi8
The desktop app updates itself from GitHub Releases. See the user guide, the changelog and the privacy policy.