# speechbrain/speechbrain

A PyTorch-based Speech Toolkit

Repository: https://github.com/speechbrain/speechbrain
Canonical: https://ross.abutalabs.com/products/speechbrain
Homepage: http://speechbrain.github.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: speech-recognition, speech-toolkit, speaker-recognition, speech-to-text, speech-enhancement, speech-separation, audio, audio-processing, speech-processing, speechrecognition, asr, voice-recognition, spoken-language-understanding, speaker-diarization, speaker-verification, pytorch, huggingface, transformers, language-model, deep-learning
Last push: 2026-08-25T22:43:42+00:00

## Health v2 (maintenance only)
Score: 83/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 53, longevity 100
- inputs: {"age_days": 2318, "days_push": 8, "days_rel": 156, "gap_med": 93.5, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 11785, forks 1718 (observed 2026-08-28T04:10:49.780325+00:00)

## What it is
SpeechBrain is an open-source PyTorch-based speech toolkit for building conversational AI systems. It provides training recipes, pretrained models, and tools for speech recognition, speaker recognition, speech enhancement/separation, TTS, and language modeling.

## Use cases
- transcribe speech to text with pretrained models
- fine-tune whisper or wav2vec2 on my own dataset
- verify speaker identity from voice recordings
- separate overlapping speakers in an audio file
- enhance noisy speech recordings
- train a speech recognition model from scratch
- build a voice assistant pipeline
- diarize who spoke when in a meeting recording

## When to choose
- you want a flexible PyTorch-based toolkit covering many speech tasks with 200+ training recipes
- you need to fine-tune HuggingFace pretrained speech models like Whisper, Wav2Vec2, or WavLM
- you are doing speech research and want customizable, well-documented training pipelines

## When to avoid
- you only need a quick off-the-shelf transcription API with no training or customization
- your project is not Python/PyTorch based
- you need production-grade low-latency streaming ASR out of the box without building it yourself

## Facets
- artifact type: library
- maturity: active
- function: speech-recognition, audio-processing, machine-learning, deep-learning, tts, nlp, llm-training
- domain: speech-processing, machine-learning, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: pytorch, conversational-ai, asr, speaker-recognition, speech-enhancement, speech-separation, huggingface, training-recipes, diarization, natural-language-processing, audio, gpu

## Member repositories
- speechbrain/speechbrain (main) score 83

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:49.780325+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 2026-08-29T17:15:19.848804+00:00, confidence not recorded.
  - readme: https://github.com/speechbrain/speechbrain (fetched 2026-08-28T04:10:49.780325+00:00, sha 031ce6dcdd09)
  - homepage: http://speechbrain.github.io (fetched 2026-08-29T08:13:22.735418+00:00, sha aaa98d8f6f82)
- Data as of 2026-08-30T08:39:29.467469+00:00.