Wireface GitHub

Get started

Running in a few minutes.

Wireface Desktop is free for Windows. The installer brings everything it needs; you bring the agent: Claude Code, OpenCode or Codex, set up the way you like it.

Read the install guide Try it in your browser

Wireface Desktop 0.1.0

Download for Windows

WirefaceSetup-0.1.0.exe · 133 MB · 64-bit · all releases on GitHub

SHA-256 61eb1b5c8060a6d32c4fad7bfbf24903837ce244affb18e61f5a236bd47c1e9b

Check it before you run it. In PowerShell, in the folder you saved it to, this prints True:

(Get-FileHash .\WirefaceSetup-0.1.0.exe -Algorithm SHA256).Hash -eq '61eb1b5c8060a6d32c4fad7bfbf24903837ce244affb18e61f5a236bd47c1e9b'

The installer isn't code-signed yet, so Windows may say "Windows protected your PC". Once the hash matches, choose More info, then Run anyway.

What you need

Before you install

Nothing else to install: Wireface brings its own Python, Electron and MediaPipe, and installs for your user, so it needs no administrator rights.

  • Windows 10 or 11, 64-bit.
  • An agent, installed and signed in: Claude Code, OpenCode or Codex, with any skills and MCP servers you want the face to have.
  • About 650 MB for the speech models, downloaded once, the first time it starts.
  • A microphone to talk to it. You can type to it too.
  • For the realtime voice, a Gemini or OpenAI API key (optional). Without one, a voice that runs on your PC reads the agent's replies.

First steps

From the download to "hello"

  1. 1

    Install it

    Run WirefaceSetup-0.1.0.exe. It installs in %LOCALAPPDATA%\Programs\Wireface and adds Wireface to the Start menu.

  2. 2

    Start it

    The first time, it downloads the speech models (about 650 MB) in a window that shows the progress. Then the face appears.

  3. 3

    Connect your agent

    Right-click the face for settings. Choose Claude Code, OpenCode or Codex, the folder it works in, and add a voice key if you have one.

  4. 4

    Say hello

    Turn on listening and talk to it. Or type, or drop a file on the face.

The install guide

Each step in more detail, with the settings worth knowing about and what to do if something doesn't go to plan.

Read the install guide

Updating and uninstalling

To update, run a newer installer over the old one. Uninstall it from Settings, then Apps: that removes the app and the speech models, and keeps your settings, conversations and logs in its data folder either way.

Other ways in

From source, or in your own app

Run Wireface Desktop from source

You need Python 3.11 or newer and Node.js. The first run sets up a Python venv, Electron and the speech models, all inside the folder.

git clone https://github.com/compsmart/wireface-desktop wireface
cd wireface
Wireface.cmd

Several faces at once? Wireface.cmd --face work gives each one its own settings, folder and voice.

Wireface Desktop on GitHub

Build the face into your own app

Wireface core is the face engine on its own, for the browser: the same talking, lip-synced 3D face, with skins, expressions and teeth. Plain ES modules on WebGL2, with no build step and no dependencies.

npm install github:compsmart/wireface-core

Then import createFace from @wireface/core, or copy the folder into your project.

Wireface core on GitHub Try it in your browser

Free for personal use

Using it at work, for clients or in something you sell? A commercial licence is £1,000 once, for your whole organisation, with no limits.

Download for Windows See pricing