# Kieirra/murmure

Fully local, private and cross platform Speech-to-Text with LLM Post-processing

Repository: https://github.com/Kieirra/murmure
Canonical: https://ross.abutalabs.com/products/murmure
Homepage: https://www.murmure.app
Language: Rust
License: AGPL-3.0
License Family: copyleft
Topics: privacy, speech-to-text, asr, asr-model, debian-packages, linux, macos, windows
Last push: 2026-09-02T15:27:42+00:00

## Health v2 (maintenance only)
Score: 85/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 100, longevity 23
- inputs: {"age_days": 325, "days_push": 0, "days_rel": 2, "gap_med": 21.5, "n_releases_24m": 13}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1047, forks 95 (observed 2026-09-03T02:15:06.399511+00:00)

## What it is
Murmure is a privacy-first, open-source desktop speech-to-text application that transcribes voice entirely on-device using NVIDIA's Parakeet TDT 0.6B v3 model, with optional LLM post-processing. It supports 25 European languages, requires no internet connection, and collects no data or telemetry.

## Use cases
- dictate text into any app instead of typing
- transcribe voice notes offline without cloud services
- dictate prompts into AI coding tools like Cursor or Claude
- fix grammar or translate dictated text with a local LLM
- write hands-free notes for privacy-sensitive work like healthcare or journalism
- use a phone as a wireless microphone for desktop dictation

## When to choose
- you need fully offline transcription with zero data leaving your machine
- you dictate in one of the 25 supported European languages
- you want open-source, auditable speech-to-text with no telemetry or accounts
- you want to post-process transcripts with your own local or remote LLM

## When to avoid
- you need transcription for languages outside the 25 supported European ones
- you need cloud-scale batch transcription of large audio archives
- you require mobile (iOS/Android) dictation rather than desktop

## Facets
- artifact type: application
- maturity: active
- function: speech-recognition, nlp, llm-inference, gui
- domain: speech-processing, privacy, developer-tools
- platform: windows, cross-platform, rust
- tags: speech-to-text, dictation, offline, parakeet, voice-typing, local-llm, no-telemetry, natural-language-processing, macos, linux, desktop

## Member repositories
- Kieirra/murmure (main) score 85

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:06.399511+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:06:37.517570+00:00, confidence not recorded.
  - readme: https://github.com/Kieirra/murmure (fetched 2026-09-03T02:15:06.399511+00:00, sha 28efaf6e235c)
  - homepage: https://www.murmure.app (fetched 2026-08-29T13:06:10.263182+00:00, sha 8effc292b34b)
  - site_page: https://docs.murmure.app/ (fetched 2026-08-29T13:06:10.272262+00:00, sha 550c1e60276a)
- Data as of 2026-08-30T08:39:29.467469+00:00.