Claude read aloud: voice mode and other ways to listen
Many people want Claude to read its answers aloud, on the phone or at a desk. Anthropic’s help center describes this as voice mode, a spoken conversation with Claude, rather than a read-aloud button for text replies.
This guide explains how voice mode works according to the Claude help center, how to choose the voice and language, and what to use when you want an existing text answer read aloud or saved as an MP3.
Voice mode: Claude speaks its replies
According to Anthropic, voice mode works in the Claude mobile apps for iOS and Android, in Claude Desktop and on the web, on all plans. On the web and desktop, tap the sound wave symbol in the lower right corner of the chat window; on mobile, tap the sound wave symbol next to the microphone in the text field. Voice conversations count toward your usual usage limits, and transcripts are saved in your chat history.
Choose the voice, pace and language
On the web and desktop, open Settings, then Capabilities, and scroll to Voice settings; if it is not there, look under Settings, General. Click a voice to preview it. On mobile, tap the settings button in the bottom left corner during voice mode to pick a voice and pace. The spoken language is set under Voice, Spoken language, and changing the app’s display language does not change it; languages other than English are in beta.
Hear a text answer read aloud
The help article describes voice mode and dictation, not a read-aloud button for written replies. To listen to an answer you already have, use your browser or system: Read Aloud in Microsoft Edge, Reading mode in Chrome, or Speak Selection on an iPhone. Our guides to Read Aloud in Edge, Word and Chrome and to the Read Aloud extension cover the shortcuts.
Save an answer as an MP3
To keep an answer for later, copy it into a text file and voice it with the free SuperClonica engine: install it with python -m pip install superclonica-cpu and run superclonica --text-file answer.txt --voice F3 --lang en --normalize --output answer.mp3. It runs offline on an ordinary processor with no character limit, in 31 languages. Hear catalog voices free on their pages first: paste 30 to 100 characters, no account needed.
For developers
If you build an app on the Claude API and want it to speak, pair the text with a speech engine. The SuperClonica engine runs on your own server, and the SuperClonica speech API takes up to 20,000 characters per request, with a free key for 3 minutes a day.