Voice Input & Dictation
Talk to Muin instead of typing — record voice messages in chat, have them transcribed automatically, and speak quick commands that turn into actions.
Anywhere you’d type to Muin’s assistant, you can speak instead. Voice input records a short clip, transcribes it automatically, and treats the words exactly as if you’d typed them — including assistant questions and quick commands.
Using Voice in Chat
- Open Chat and find the microphone button next to the message box.
- Click to start recording — a pulsing indicator with a timer shows you’re live.
- Click stop when you’re done (or discard with Esc if you fumbled it).
- The transcription lands in the conversation and the assistant responds as usual.
Speak naturally: “What’s our total giving this month compared to last?” works just as well spoken as typed.
Voice Commands
Short, action-shaped utterances are recognized as commands rather than chat — for example dictating an expense or asking to find an invoice. Muin auto-detects whether you’re commanding or conversing, so you don’t have to switch modes.
What’s Supported
- Audio formats: recordings are handled in common formats (WAV, MP3, WebM, OGG, FLAC, M4A) — the in-app recorder picks the right one for your browser automatically.
- Languages: transcription supports multiple languages, not just English. Speak in the language you work in.
- Length: keep clips conversational — a sentence or two transcribes fastest and most accurately.
Accuracy Tips
- Reduce background noise where you can; a quiet room beats a busy lobby.
- Names and organization-specific terms transcribe best when spoken clearly — you can always edit the transcription before acting on the result.
- Per-user and per-organization usage limits protect against runaway costs; if you hit one, wait a minute and try again.
Privacy
Voice clips are processed only to produce your transcription. As with all Muin AI features, the results stay scoped to your organization.