---
title: Use Voice mode
sidebar_position: 13
---

# Use Voice mode

Voice mode lets you start a task in Work using your voice instead of typing. Your speech is transcribed into text that you can review, edit, and send as a prompt. It's available on the web app and mobile apps ([iOS](https://apps.apple.com/us/app/le-chat-by-mistral-ai/id6740410176) and [Android](https://play.google.com/store/apps/details?id=ai.mistral.chat)) for all plans.

:::info
Voice mode is powered by , a speech-to-text model we developed for fast, accurate transcription across languages.
:::

<SectionTab as="h2" sectionId="how-to-use">How to use</SectionTab>

1. Click the microphone icon at the bottom right of the chat window.
2. Allow microphone access when prompted by your browser or device.
3. Speak naturally.

What happens next depends on the `Autosend` setting:

- **Autosend ON**: your speech is transcribed and sent to Work automatically. No manual step needed.
- **Autosend OFF**: the transcribed text appears in the message box. You can review it, correct any errors, attach context (Connectors, Libraries, Skills, files), and click `Send` when you're ready.

To discard a recording, click the cancel icon before sending.

:::tip
If you previously denied microphone access, enable it in your browser's site settings or your device's app settings.
:::

<SectionTab as="h2" sectionId="conversation-mode">Talk with Work in conversation mode</SectionTab>

<SectionTab as="h2" sectionId="supported-languages">Supported languages</SectionTab>

- English
- French
- Spanish
- German
- Russian
- Chinese
- Japanese
- Italian
- Portuguese
- Dutch
- Arabic
- Hindi
- Korean

<SectionTab as="h2" sectionId="privacy">Privacy</SectionTab>

How we handle voice data depends on your plan:

- **Free, Pro, and Student plans**: voice messages fall under our data training policies and may be used to improve our models. You can [opt out](https://help.mistral.ai/en/articles/455207-can-i-opt-out-of-my-input-or-output-data-being-used-for-training) at any time.
- **Team and Enterprise plans**: voice messages aren't used for model training by default. Your organization would need to explicitly opt in.

<SectionTab as="h2" sectionId="troubleshooting">Troubleshooting</SectionTab>

If your voice message wasn't transcribed correctly:

- **Speak clearly** and reduce background noise.
- **Check microphone permissions** in your browser's site settings or your device's app settings.
- **Review and edit** the transcription in the message box before sending — especially for exact names, links, file paths, or numbers.