Automatic speaker diarization
A multi-person recording is split by voice so the transcript reads like a conversation instead of one wall of text. Each speaker gets a clear, color-coded label.
Speaker identification
Speaker identification that tells you who said what
Murmur does more than speech to text. It performs speaker diarization — separating a multi-person recording by voice — labels every transcript line, and remembers people you name so they can be recognized in later recordings, all on your device.
Recognize the same person next time
Name a speaker once and Murmur keeps that voice in your local people library. When that person appears in another recording, Murmur can identify them automatically.
Correct mistakes without losing control
Rename speakers, merge duplicates, relearn a voice, or reassign an individual line. The speaker data stays on the device together with your recordings and transcripts.
Use cases
Meeting transcription
Put the phone on the table, hit record, and stay in the conversation. Murmur writes the meeting down as it happens and labels every line with the person who said it — on your device, without sending the room to anyone else.
Interview transcription
Journalists, researchers and anyone recording a conversation they promised to protect: Murmur produces a verbatim, speaker-separated transcript without a single byte leaving the phone.
Audio file transcription
A voice memo, a meeting somebody else recorded, a lecture, a video file. Import it into Murmur and get a transcript on your own device, with every speaker labeled and every line replayable.
Common questions
What is speaker diarization?
Speaker diarization answers “who spoke when” by separating a recording into different speakers. Murmur combines diarization with speaker labels and local speaker recognition.
Is speaker identification processed in the cloud?
No. Speaker separation, recognition and the people library all run and stay on your device.
Can it remember speakers between meetings?
Yes. Once you name a voice, Murmur can recognize that person in later recordings.
Start with your next conversation.
Download Murmur, download a speech model once, and every recording after that is transcribed on your own device — meetings, interviews, lectures, ideas on a walk.
Free to download · No account · Works with no signal
On iPhone and iPad the free version transcribes the first 20 minutes of each recording; Murmur Pro is a one-time purchase that lifts the limit. On Android, transcription is currently free with no recording-length limit.