streaming speech-to-text / real-time ASR with diarization
Real-time, self-hosted speech-to-text pipeline with streaming ASR, speaker diarization, simultaneous translation, and OpenAI/Deepgram-compatible APIs.
Same problem, same approach. Swapping one for another is a config change, not a rewrite.
when you need a trainable ASR toolkit with VAD/punctuation and OpenAI-compatible serving, especially for Chinese
Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
when deploying on embedded/CPU targets with C++ and ONNX runtime instead of a Python/PyTorch server
Speech-to-text, text-to-speech, speaker diarization, speech enhancement, source separation, and VAD using next-gen Kaldi with onnxruntime without Internet connection. Support embedded systems, Android, iOS, HarmonyOS, Raspberry Pi, RISC-V, RK NPU, Axera NPU, Ascend NPU, x86_64 servers, websocket server/client, support 12 programming languages
when you need full control to train/customize speech models rather than a ready pipeline
A PyTorch-based Speech Toolkit
when you need ultra-low-latency on-device voice agent STT in C++ and can skip diarization/translation
Very low latency speech to text, intent recognition, and text to speech, for building voice agents and interfaces
when building Apple apps and want CoreML-based on-device STT/diarization in Swift
Frontier CoreML audio models in your apps — text-to-speech, speech-to-text, voice activity detection, and speaker diarization. In Swift, powered by SOTA open source.
when building Apple Silicon apps with MLX/CoreML and need a Swift speech toolkit
AI speech toolkit for Apple Silicon — ASR, TTS, speech-to-speech, VAD, and diarization powered by MLX and CoreML
when you need very fast CPU streaming STT in Rust with NVIDIA Parakeet and no server features
very fast speech-to-text, diarization, streaming (even in CPU) with NVIDIA Parakeet in Rust
when building Apple apps with on-device speech models in Swift; similar to FluidAudio but different model set
On-device Speech AI for Apple Silicon
when you want a production streaming ASR toolkit you can train/deploy; no built-in diarization/translation
Production First and Production Ready End-to-End Speech Recognition Toolkit
when you prefer PaddlePaddle toolkit covering ASR/TTS/translation; different framework
Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.
when you need a research speech-processing toolkit to build custom pipelines; not ready-made
End-to-End Speech Processing Toolkit
when you need lightweight offline streaming ASR with bindings for many languages; no diarization/translation
Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
when you want a simple offline Whisper+pyannote diarization notebook; no streaming
Automatic Speech Recognition with Speaker Diarization based on OpenAI Whisper
when you need only a batch Whisper REST API without streaming or diarization
OpenAI Whisper ASR Webservice API
when you need accurate word-level timestamps/confidence from Whisper offline
Multilingual Automatic Speech Recognition with word-level timestamps and confidence
when targeting embedded CPUs with ncnn and need no diarization
Real-time speech recognition and voice activity detection (VAD) using next-gen Kaldi with ncnn without Internet connection. Support iOS, Android, Linux, macOS, Windows, Raspberry Pi, VisionFive2, LicheePi4A etc.
when you need a minimal offline Whisper+pyannote diarization glue; no streaming/translation
when you need offline transcripts with speaker names/recognition; no streaming/translation
Speechlib is a library that unifies speaker diarization, transcription and speaker recognition in a single pipeline to create transcripts for audio conversations with actual speaker names and time tags. This library also contains audio preprocessor functions.
when you want a minimal streaming Whisper library to build your own pipeline; no diarization/API
Whisper realtime streaming for long speech-to-text transcription and translation
when you need fast offline word-level timestamps and diarization from Whisper; no streaming
WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)
when you need a simple multi-engine Python wrapper for basic ASR; lacks streaming/diarization
Speech recognition module for Python, supporting several engines and APIs, online and offline.
when you need an older lightweight offline STT engine/training with MPL license; stale
🐸STT - The deep learning toolkit for Speech-to-Text. Training and deploying STT models has never been so easy.
Does the same job, but ships as an app. Useful to a person, not swappable into a codebase.
when you want a ready local STT app/CLI with LLM post-processing rather than a server to integrate
Fully local, private and cross platform Speech-to-Text with LLM Post-processing
when you want to run Whisper as a standalone Windows executable without managing Python
Whisper & Faster-Whisper standalone executables for those who don't want to bother with Python.
These projects were analysed and named WhisperLiveKit among their alternatives. The relationship is not symmetric — how WhisperLiveKit rates them is a separate judgement, made when WhisperLiveKit is analysed in its own right.
calls WhisperLiveKit “choose for real-time streaming ASR with diarization and OpenAI/Deepgram-compatible APIs”
WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)
calls WhisperLiveKit “real-time STT server with streaming and diarization, OpenAI-compatible API.”
Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node
calls WhisperLiveKit “Choose when you want Whisper-specific streaming with diarization/translation; it is an alternative library but not API-compatible.”
A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.
calls WhisperLiveKit “choose for real-time streaming ASR with diarization and translation.”
Speech recognition module for Python, supporting several engines and APIs, online and offline.
calls WhisperLiveKit “Choose when you need real-time local streaming ASR with diarization and OpenAI/Deepgram-compatible APIs.”
ML-powered speech recognition directly in your browser
calls WhisperLiveKit “choose for a ready-made real-time local ASR server with OpenAI-compatible APIs, based on Whisper.”
Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
calls WhisperLiveKit “choose when you need a local real-time STT server with OpenAI/Deepgram-compatible APIs.”
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
calls WhisperLiveKit “An API server/tool that wraps Whisper-style streaming ASR and diarization rather than an embeddable SDK.”
Speech-to-text, text-to-speech, speaker diarization, speech enhancement, source separation, and VAD using next-gen Kaldi with onnxruntime without Internet connection. Support embedded systems, Android, iOS, HarmonyOS, Raspberry Pi, RISC-V, RK NPU, Axera NPU, Ascend NPU, x86_64 servers, websocket server/client, support 12 programming languages