wenet-e2e/wenet

Production First and Production Ready End-to-End Speech Recognition Toolkit

What it solves

WeNet is a production-ready toolkit for end-to-end speech recognition. It aims to bridge the gap between research and deployment by providing a full-stack solution that is accurate, lightweight, and designed for real-world production environments.

How it works

WeNet provides a framework for both streaming and non-streaming speech recognition. It supports multiple models such as Paraformer, FireRed, and Whisper (large-v3 and large-v3-turbo) for English and Chinese transcription. The system includes a Python package for easy transcription and a separate runtime for deployment, which can be built for x86 platforms to optimize performance.

Who it’s for

Developers and engineers who need to implement high-accuracy speech-to-text systems in production, as well as researchers who want to build and train end-to-end speech recognition models.

Highlights

  • Production-ready: Designed specifically for production environments with full-stack solutions.
  • SOTA results: Achieves state-of-the-art results on various public speech datasets.
  • Easy Installation: Available as a Python package for quick start or a full installation for training and deployment.
  • Broad Hardware Support: Supports CUDA and Ascend NPU for acceleration.

Related

  • Project
    espnet/espnet

    An end-to-end speech processing toolkit that provides a unified framework for ASR, TTS, speech translation, and enhancement using PyTorch.

  • Project
    wenet-e2e/WeTextProcessing

    A production-ready text processing toolkit for text normalization and inverse text normalization, supporting multiple languages for speech-related AI applications.

  • Project
    PaddlePaddle/PaddleSpeech

    An open-source speech and audio toolkit on the PaddlePaddle platform providing state-of-the-art models for speech recognition, synthesis, and translation.

  • Project
    modelscope/FunASR

    An industrial speech recognition toolkit for offline, streaming, and edge deployment, providing high-performance ASR, VAD, and speaker diarization pipelines.

  • Project
    QuentinFuxa/WhisperLiveKit

    An ultra-low-latency, self-hosted speech-to-text pipeline that enables real-time transcription and translation using advanced simultaneous speech research and multiple model backends.