Skip to main content
Install SpeakoFlow, then open it. A short setup runs on first launch.

Onboarding

Permissions

On macOS the app asks for Microphone and Accessibility permission. This screen skips itself if you have already granted them, and it does not appear on Windows or Linux.See Troubleshooting if a permission gets stuck after an update.

Pick a speech model

Five models are offered. Each card shows the model name, the languages it hears, one sentence about what it is for, and its download size (or that it is already installed).Parakeet and Nemotron carry a “Recommended” badge. The model for your language is preselected. “Set up later” skips the download and lets you choose on the Models page instead.The download runs in the background, so you do not wait. The full catalog (65 models) is on the Models page.

Tour

Three scenes show what SpeakoFlow does, each with the real surface you will see beside the steps:
  1. Dictation. Hold the keys, talk, let go. Your words are typed at the cursor.
  2. Assistant. Select text, hold the keys, and ask. Translate, reply, explain.
  3. Conversation. Press the keys and talk. The assistant answers out loud.
The tour only shows. It downloads nothing and switches nothing on.
The tour’s last button opens the app. If the speech model is still downloading, the sidebar shows progress, and a dictation pressed before it finishes shows a “downloading” pill in the overlay instead of recording.

Dictate your first sentence

Open any text field, hold your Dictate shortcut, and talk. Let go, and your words type themselves. Every shortcut is changeable. See Keyboard shortcuts for the full table. Next: keyboard shortcuts.
Last modified on August 30, 2026