# ufal/whisper_streaming

Whisper realtime streaming for long speech-to-text transcription and translation

Repository: https://github.com/ufal/whisper_streaming
Canonical: https://ross.abutalabs.com/products/whisper_streaming
Language: Python
License: MIT
License Family: permissive
Last push: 2025-11-12T16:45:41+00:00

## Health v2 (maintenance only)
Score: 53/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 51, release rhythm 35, longevity 89
- inputs: {"age_days": 1246, "days_push": 294, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3672, forks 408 (observed 2026-08-28T04:08:13.566531+00:00)

## What it is
A Python library that turns Whisper-like speech recognition models into a real-time streaming transcription and translation system using a local agreement policy with self-adaptive latency. It supports multiple Whisper backends such as faster-whisper and whisper-timestamped and was demonstrated in live multilingual conference transcription.

## Use cases
- transcribe live audio streams in real time
- real-time speech-to-text for long unsegmented audio
- live captioning for conferences and meetings
- streaming speech translation across languages
- build a live transcription service on top of Whisper

## When to choose
- you need low-latency streaming transcription from Whisper models
- you want multilingual live captioning or translation
- you want to plug in different Whisper backends like faster-whisper

## When to avoid
- you need the latest state-of-the-art simultaneous speech recognition - the project recommends its successor SimulStreaming
- you only need offline batch transcription of pre-recorded files
- you need a production-ready managed service rather than a research-grade library

## Facets
- artifact type: library
- maturity: maintenance
- function: speech-recognition, streaming, nlp, audio-processing
- domain: speech-processing, machine-learning
- platform: python, cross-platform
- tags: whisper, real-time-transcription, translation, local-agreement, simultaneous-translation, natural-language-processing, gpu

## Member repositories
- ufal/whisper_streaming (main) score 53

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:13.566531+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 2026-08-29T18:31:18.541218+00:00, confidence not recorded.
  - readme: https://github.com/ufal/whisper_streaming (fetched 2026-08-28T04:08:13.566531+00:00, sha 23e94757e731)
- Data as of 2026-08-30T08:39:29.467469+00:00.