# OpenMOSS/MOSS-Transcribe-Diarize

MOSS-Transcribe-Diarize 0.9B is an open-source SOTA end-to-end audio understanding model for long-form multi-speaker transcription, diarization, timestamps, and acoustic event awareness.

Repository: https://github.com/OpenMOSS/MOSS-Transcribe-Diarize
Canonical: https://ross.abutalabs.com/products/moss-transcribe-diarize
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-26T15:27:22+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 8
- inputs: {"age_days": 112, "days_push": 7, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1695, forks 98 (observed 2026-08-28T04:05:23.540814+00:00)

## What it is
MOSS-Transcribe-Diarize 0.9B is an open-source end-to-end audio understanding model that jointly performs multi-speaker speech transcription and speaker diarization with timestamps and acoustic event annotations. It supports 50+ languages and includes serving options (SGLang, vLLM) and a subtitle-generating web UI.

## Use cases
- transcribe long multi-speaker meetings with speaker labels
- diarize podcasts and interviews with timestamps
- generate subtitles from video or lecture recordings
- transcribe multilingual audio in 50+ languages
- detect acoustic events in recordings alongside transcripts
- transcribe phone calls with consistent speaker labels

## When to choose
- you need joint ASR and diarization in a single pass instead of stitching separate systems
- you work with long, messy, multi-speaker recordings like meetings, calls, or podcasts
- you need open-source, self-hosted transcription with timestamps and speaker labels
- you need multilingual transcription across 50+ languages

## When to avoid
- you only need simple single-speaker dictation where a lightweight ASR suffices
- you have no GPU or inference infrastructure available
- you need a fully managed hosted service rather than a self-run model
- you need real-time streaming transcription rather than long-form batch processing

## Facets
- artifact type: library
- maturity: active
- function: speech-recognition, machine-learning, llm-inference, audio-processing
- domain: speech-processing, machine-learning
- platform: python
- tags: asr, diarization, speaker-labeling, timestamps, transcription, multilingual, subtitles, long-form-audio, natural-language-processing, audio, gpu, linux, docker

## Member repositories
- OpenMOSS/MOSS-Transcribe-Diarize (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:23.540814+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 2026-08-30T03:37:55.350485+00:00, confidence not recorded.
  - readme: https://github.com/OpenMOSS/MOSS-Transcribe-Diarize (fetched 2026-08-28T04:05:23.540814+00:00, sha fca21dcd3182)
- Data as of 2026-08-30T08:39:29.467469+00:00.