About Cue

Voice AI that works where you work.

Cue is a desktop voice AI for Mac and Windows. Press a hotkey, speak, and Cue can transcribe, use the screen context in front of you, and help finish the task in the app you are already using.

What Cue is

Cue is built for people who want to move through desktop work by voice instead of moving every task into a separate chat tab. It supports macOS 13+ and Windows 10+ (x64). Cue is free to start, and Cue Plus is $19.99/month.

The product combines voice typing with screen-aware desktop workflows. The user remains responsible for reviewing AI-generated output and actions, especially when a mistake could have a meaningful consequence.

Who makes Cue

Cue is developed by Sophon LLC. This website is the canonical public source for Cue product information, downloads, pricing, privacy terms, and published product evidence.

For product support, account or billing questions, privacy requests, security concerns, or general feedback, use the Cue contact page.

Verify the official Cue website

Cue is the desktop voice agent developed by Sophon LLC. Its official website is https://heycue.io. Use these checks when you arrive from search, a shared link, or a download recommendation.

  1. Check the website: look for heycue.io in your browser's address bar. Compare the full domain rather than relying on a page title, product name or logo.
  2. Start downloads here: open the official download section and choose the installer for your computer. Follow the download links published there; the installer itself may be delivered from a separate file host.
  3. Find support here: open the official Contact page on this domain for the current support and privacy email. That page also explains what to include in a useful request.

The Privacy Policy and Terms of Service identify Sophon LLC and describe the service. If you are unsure about a link, return to this site directly before downloading or sharing account information.

Published evidence

Cue is Featured in Google DeepMind's Gemmaverse. The published May 2026 case study reports 44% lower median polish latency for the measured on-device Gemma 4 text-polish path, a 30% increase in feature usage, and zero marginal inference cost for that local polish step.

Read the Google DeepMind case study or see Cue's evidence page.

Accountability

Cue publishes its current Privacy Policy and Terms of Service. Those pages describe data handling, third-party services, subscriptions, acceptable use, and the responsibilities that come with AI-generated content and actions.