Talk to Your Agent — Voice Input Lands in Sidekick v1.0.60

Released in v1.0.60

Not everything is faster to type. Sometimes you're on the go, your hands are full, or you just think better out loud. Starting today, you can press and hold to speak directly to your Sidekick agent — no keyboard required.

How It Works

Press and hold the microphone button in the chat interface, say what you need, and let go. Sidekick transcribes your recording on the server side and passes your words to the agent exactly as if you'd typed them. The agent's reply appears in the same thread, ready for you to read or act on.

There's nothing to set up. No third-party transcription service to connect, no API key to paste in, and no app to install. It works in the browser, on desktop or mobile, the moment you load the dashboard.

Built for Mobile

Voice input is especially useful on mobile. Typing a detailed instruction on a small screen takes time and attention. Speaking it takes seconds. Whether you're asking your agent to reschedule a meeting, summarise your unread emails, or file a GitHub issue, voice gets you there faster when you're away from your desk.

Telegram and WhatsApp Voice Notes

Maybe you already have a habit of sending yourself voice notes in Telegram or WhatsApp. Now those notes can go straight to your agent. Forward any voice note to your connected Sidekick agent and it'll be transcribed and processed just like any other message. Capture a thought on a walk, forward it, and come back to find the agent has already acted on it.

What's Next

This release focuses on voice input — getting your words into the agent quickly and naturally. If you have feedback on transcription quality or want to see voice output (agent responses read aloud), let us know through the in-app feedback button.

Full details are in the changelog and the Voice Input feature guide.