Your first voice in 60 seconds
This guide takes you from a fresh install to hearing VoxStudio speak.
1. Launch VoxStudio
Open the app. A short Warming up VoxStudio screen appears while the voice engine starts. The ring of bars sweeps while it works, and each step is ticked off as it finishes:
- Launching voice engine: VoxStudio starts voxd in the background.
- Preparing runtime: on the very first launch, VoxStudio sets up its private Python runtime. This takes a minute or two and happens only once.
- Checking voices: voxd looks for the voice engines available on your computer.
When everything is ready, the screen fades away. The sidebar footer shows a green dot and Engine ready.
First launch: pick your voices
The first time, a short welcome sheet appears:
- Welcome: what makes VoxStudio different.
- Choose your first voices: Kokoro (recommended: 54 natural voices, 354 MB) or just system voices (no download).
- Download: watch it progress, or choose Continue in background and keep exploring. When it's done, Start creating switches Home to your new voices.
If it doesn't start: click Show details to see the engine log, then Try again. The most common cause is a missing
uv. Install it from https://docs.astral.sh/uv/ and relaunch.
2. Speak
On Home, find the Try a voice card.
- Edit the text, or keep the sample sentence.
- If Kokoro is installed, choose System or Kokoro in the top-right of the card.
- Pick a voice. Voices are grouped by language, with yours first; ♀ and ♂ mark the voice's gender.
- Choose a pace: Calm, Natural or Brisk.
- Click Speak, or press ⌘↩.
The orb reacts to the audio as it plays.
3. Replay
Every generation is saved as a take. Your latest takes appear under the text box; click ▶ to replay one.
Next steps
- Search anything with ⌘K.
- Switch between light and dark in Settings → Appearance.
- Script it: API quickstart.