Instead of typing your prompt, you can speak it. Caffeine includes a voice input button in the chat that records what you say, transcribes it, and places the text into the chat input — ready to send or edit before you submit.
How to use it
Click the microphone icon in the chat input bar. Speak your prompt clearly, then stop. The recording is transcribed automatically and the text appears in the chat input field. Review it, make any corrections, then send as normal.
When it's useful
Voice input is particularly useful for longer, more descriptive prompts — the kind where typing everything out would take time. Describing a complex layout, a multi-step workflow, or a detailed feature is often faster to speak than to type.
Frequently asked questions
What languages are supported?
Voice input works in any language. The transcription service handles the language automatically based on what you speak.
Is the transcription always accurate?
It is generally accurate for clear speech, but technical terms, proper nouns, and unusual phrasing may occasionally be misheard. Always review the transcribed text before sending — you can edit it in the chat input field before submitting.
What happens if the transcription fails?
If the transcription cannot be completed — for example due to a network issue or a temporary service problem — Caffeine will show an error message. Your recording is not submitted, and you can try again.
Does voice input work on mobile?
Yes, on devices with a microphone and a supported browser.
Is my voice recording stored?
The audio is sent to the transcription service to produce text and is not retained after transcription.