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 provide a full-stack solution that bridges the gap between research-grade speech-to-text models and actual production environments, ensuring high accuracy and lightweight deployment.

How it works

WeNet supports both streaming and non-streaming speech recognition. It provides a variety of pre-trained models (such as Paraformer and FireRed) for different languages, including Chinese and English. The system allows users to transcribe audio files via a simple Python API or command-line interface, and it includes a dedicated runtime for optimized deployment on various platforms (including x86 and Ascend NPU).

Who it’s for

This toolkit is designed for developers and engineers who need to implement high-performance speech recognition in real-world applications, as well as researchers who want to benchmark their models against state-of-the-art results on public datasets.

Highlights

  • Production-Ready: Focused on full-stack production solutions for speech recognition.
  • SOTA Accuracy: Achieves state-of-the-art results on numerous public speech datasets.
  • Lightweight: Easy to install and well-documented for rapid deployment.
  • Hardware Support: Supports CPU, GPU (CUDA), and Ascend NPU hardware acceleration.

Related

  • 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
    espnet/espnet

    An end-to-end speech processing toolkit built on PyTorch that provides unified recipes and pretrained models for ASR, TTS, translation, and other audio AI tasks.

  • 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.

  • Project
    modelscope/FunASR

    FunASR is an industrial-grade speech recognition toolkit from Alibaba's ModelScope team. It provides a modular Python API, CLI, and server for offline/streaming ASR, plus VAD, punctuation, speaker diarization, emotion recognition, and edge deployment via llama.cpp. It includes multiple pre-trained models (e.g., Fun-ASR-Nano, SenseVoiceSmall) and supports GPU, CPU, and edge devices.