# KoljaB/RealtimeSTT

A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.

Repository: https://github.com/KoljaB/RealtimeSTT
Canonical: https://ross.abutalabs.com/products/realtimestt
Language: Python
License: MIT
License Family: permissive
Topics: python, realtime, speech-to-text
Last push: 2026-08-20T23:50:54+00:00

## Health v2 (maintenance only)
Score: 94/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 98, longevity 78
- inputs: {"age_days": 1100, "days_push": 13, "days_rel": 13, "gap_med": 8, "n_releases_24m": 26}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 10080, forks 851 (observed 2026-08-28T04:10:39.494684+00:00)

## What it is
RealtimeSTT is a Python library for low-latency, real-time speech-to-text with voice activity detection, wake word activation, and streaming transcription. It supports multiple ASR engines including faster-whisper, sherpa-onnx, and kroko-onnx for local speech recognition.

## Use cases
- transcribe microphone audio in real time
- build a voice assistant with wake word activation
- add dictation to an app
- stream speech-to-text over websocket
- detect voice activity before transcribing
- run local speech recognition on CPU or GPU

## When to choose
- you need low-latency realtime transcription in Python
- you want local, offline speech recognition with VAD and wake words
- you need streaming partial transcripts plus a final transcript

## When to avoid
- you only need batch transcription of audio files with no latency requirements
- you need a hosted cloud ASR API rather than a local library
- your project is not Python-based

## Facets
- artifact type: library
- maturity: active
- function: speech-recognition, audio-processing, streaming
- domain: speech-processing, artificial-intelligence
- platform: python, cross-platform
- tags: speech-to-text, voice-activity-detection, wake-word, realtime-transcription, whisper, microphone, voice-assistant, audio, gpu

## Member repositories
- KoljaB/RealtimeSTT (main) score 94

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:39.494684+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 2026-08-29T17:19:45.403922+00:00, confidence not recorded.
  - readme: https://github.com/KoljaB/RealtimeSTT (fetched 2026-08-28T04:10:39.494684+00:00, sha 2db639b3b7a5)
  - registry_pypi: https://pypi.org/pypi/realtimestt/json (fetched 2026-08-29T08:19:30.661210+00:00, sha 8300296d2bbd)
- Data as of 2026-08-30T08:39:29.467469+00:00.