Can Jarvis handle voice dictation and prompts?

Yes, for spoken prompts: Jarvis (getjarvis.eu) is a desktop AI assistant with a voice mode you can talk to. Its floating bar opens on Cmd+/ on Mac or Ctrl+/ on Windows, and voice mode handles questions and instructions rather than system-wide dictation into any text field. Wispr Flow and Superwhisper are the dictation-first tools if you need text typed everywhere. It is useful for hands-free questions, dictating prompts, or driving Jarvis from a meeting where typing is awkward; its free tier includes voice input. Audio travels through the OpenAI Whisper API, is encrypted in transit, and is not retained for training. Scroll down for voice-input details and download options.

Open the Jarvis floating bar with Cmd+/ on Mac or Ctrl+/ on Windows. Click the microphone icon, or hold the voice hotkey (configurable in Settings → Keyboard). Speak naturally. When you stop, Jarvis sends the audio to OpenAI's Whisper Large v3 model via API, which returns a transcript in under a second for short clips, 2-3 seconds for longer ones. The transcript appears in the prompt box. You can edit it, then submit to whichever model Jarvis routes to (a frontier model by default). For follow-up questions in a conversation, voice continues to work — you can have a fully voice-driven session if you want. The transcription is server-side; the audio file is wiped from Jarvis infrastructure after the response and not retained for training by OpenAI under the API enterprise terms.

Three things to know. (1) Jarvis is not a system-wide dictation tool — it transcribes into its own prompt box, not into the document, email, or message field your cursor is in. If you want to dictate a paragraph into Apple Notes or a Slack message, you need a tool that targets the focused text field. (2) Jarvis does not speak responses out loud (TTS) by default; voice is input-only. Read-aloud is on the roadmap but not yet shipped. (3) No always-on listening — voice input requires an explicit click or hotkey press. Privacy-first by design. If you want hands-free always-on like Cluely or Highlight, those tools fit better. If you want explicit press-to-talk for questions and prompts, Jarvis voice input is exactly that.

Capabilities