Install it on your computer and use it live.

Two builds are available: macOS (Apple silicon, macOS 13 or later) and Windows (Windows 10 version 2004 or later, or Windows 11). Once installed, you can get answers you can say out loud while the meeting is under way. After you sign in to the customer center, a signed download link valid for 5 minutes is generated; activation and download are bound to your device.

Reading the current online version

Checking the installer status.

MAC / WINDOWS
Customer center

Sign in to the customer center first.

The download button carries your download intent. If you are not signed in, finish signing in or registering first; switching language or a form error does not lose it.

Signed link

Then generate a 5-minute signed link.

After authentication succeeds, the customer center generates a signed download link valid for 5 minutes, with a traceable watermark where it can be applied. The link is verified by the server, and the public web page does not expose the installer address directly.

First run

On Windows, get past SmartScreen on first run.

The Windows build is an unsigned portable ZIP that needs no installation: unzip it and run it. On the first run Windows may show a SmartScreen prompt: choose "More info", then "Run anyway" to continue; you can also unblock the file in its properties.

System permissions

Grant system permissions as needed.

Permissions such as the microphone, system audio and screen recording are turned on only when the matching feature needs them. Transcription runs on your device by default (Apple silicon and 64-bit Windows); when you switch to the cloud fast or standard mode, audio is transcribed through the gateway, and answer generation goes through the gateway. If you hit a permission or audio source problem, check your current settings on the help page first.

Controlled download path

A controlled entry, with no anonymous direct links

The public web page, customer center authentication, a 5-minute signed download link, and the installer for your platform. If any step is unavailable, the page tells you the current status instead of turning into an anonymous direct link.

Permissions

Permissions are confirmed feature by feature.

The microphone, system audio and screen recording each take a different input role. Screen reading can be turned off on its own, and the matching permission is handled only when the feature needs it.

MicrophoneHandles the voice input you allow
System audioCan be combined with the microphone as you set
Screen recordingNeeded only when screen context is enabled

Version and status, shown as they are

The page shows the real online version; if the installer is temporarily unavailable, your account and download intent are kept, and you can try again later.