Murmur Recorder: Voice to Text

Murmur is the recorder that understands who said what. Record a meeting, a lecture, an interview, or a quick voice memo, and Murmur turns it into searchable text — and tells each speaker apart — entirely on your device. Your recordings are never sent to our servers.

WHY MURMUR

Most transcription apps upload your recordings to a server. Murmur does not. Everything happens on-device, using a speech recognition model that runs locally on your phone. No account. No sign-up. We never receive your recordings or your transcripts.

KNOW WHO SAID WHAT
Murmur labels each line of the transcript with who said it, color-coded so a multi-person conversation is easy to follow. It builds a library of people across your recordings: name someone once, and Murmur recognizes their voice the next time they appear. Merge duplicates or relearn a voice — the more you correct someone, the better Murmur identifies them later.

TURNS TALK INTO TEXT
Hit record, and Murmur transcribes what was said and saves the full transcript alongside the audio. Transcription runs on your phone and detects the spoken language automatically.

EXPORT — AND HAND IT TO YOUR AI
Export any transcript as plain text, Markdown, PDF, or Word — with each line's speaker, and your translation if you made one. Export subtitles as SRT or VTT, or share the original audio. Drop it into your notes, an email, or your favorite AI assistant to summarize, pull action items, or dig deeper — on your terms, with your tools.

IMPORT WHAT YOU ALREADY HAVE
Bring in a voice memo, a meeting recording, a lecture, or any audio or video file, and Murmur transcribes it on-device with speakers labeled — just like something you recorded in the app.

TRANSLATE YOUR TRANSCRIPTS
Turn any transcript into another language, line by line, right inside the recording. Read it as translation-only or side-by-side with the original — all on-device.

WORKS OFFLINE
Because recognition runs on your device, Murmur keeps working on a plane, in a basement meeting room, or anywhere with no signal. Download a model once and you are set.

GREAT FOR LEARNING
Import a speech, a lecture, or any audio you already have, and study it your way: tap a single sentence to replay it, or let the whole recording play with the transcript scrolling along in sync. Many people use Murmur to learn a new language one sentence at a time.

PLAY BACK WHILE YOU READ
Open any recording to read the transcript with timestamps, scrub through the audio, and replay any moment. Tap any line to hear just that sentence.

SEARCH YOUR WORDS
Find a phrase across your recordings instantly, with matches highlighted right where they appear.

MULTIPLE LANGUAGES
Murmur transcribes around 100 spoken languages — accuracy is highest for the most widely spoken ones — and the app interface is available in English, Japanese, Simplified Chinese, Korean, and French.

SAVES SPACE AUTOMATICALLY
Once a recording is transcribed, Murmur quietly compresses its audio in the background, reclaiming storage without losing the transcript or playback.

PRIVATE BY DESIGN
- 100% on-device transcription and speaker identification
- No account, no sign-up
- Your recordings and transcripts are never sent to our servers
- Anonymous usage and crash diagnostics help us fix bugs — never your audio, never your words

Whether you are capturing decisions from a meeting, keeping a record of an interview, or jotting down ideas on a walk, Murmur gives you a clean, searchable transcript that knows who said what — while keeping your conversations yours.

Download a model, hit record, and start talking.