[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-71002":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":30,"readmeContent":31,"aiSummary":32,"trendingCount":16,"starSnapshotCount":16,"syncStatus":33,"lastSyncTime":34,"discoverSource":35},71002,"RealtimeSTT","KoljaB\u002FRealtimeSTT","KoljaB","A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.","",null,"Python",9885,836,73,124,0,11,40,116,33,114.77,"MIT License",false,"master",true,[27,28,29],"python","realtime","speech-to-text","2026-06-12 04:00:58","# RealtimeSTT\n\nRealtimeSTT is a Python speech-to-text library for applications that need\nvoice activity detection, fast transcription, optional realtime text updates,\nwake words, and direct access to audio streams. It is designed for assistants,\ndictation tools, browser streaming servers, and prototypes that need to turn\nspeech into text with only a few lines of code.\n\nThe recommended default path uses `faster_whisper`. Other engines are available\nthrough install extras when their optional dependencies and models are present.\n\n## Install\n\n```bash\npip install \"RealtimeSTT[faster-whisper]\"\n```\n\nOn Linux, install PortAudio headers before installing the package:\n\n```bash\nsudo apt-get update\nsudo apt-get install python3-dev portaudio19-dev\n```\n\nOn macOS:\n\n```bash\nbrew install portaudio\n```\n\nFor CUDA, platform notes, and optional engine stacks, see\n[docs\u002Finstallation.md](docs\u002Finstallation.md).\n\n## Microphone Example\n\nThis waits for speech, stops after the detected utterance, and prints the final\ntranscript:\n\n```python\nfrom RealtimeSTT import AudioToTextRecorder\n\nif __name__ == \"__main__\":\n    with AudioToTextRecorder() as recorder:\n        print(\"Speak now\")\n        print(recorder.text())\n```\n\nUse the `if __name__ == \"__main__\":` guard when running scripts, especially on\nWindows, because RealtimeSTT uses multiprocessing for model work.\n\n## Automatic Recording Loop\n\nFor continuous dictation, pass a callback to `text()` so transcription work can\ncomplete asynchronously while your loop keeps listening:\n\n```python\nfrom RealtimeSTT import AudioToTextRecorder\n\n\ndef process_text(text):\n    print(text)\n\n\nif __name__ == \"__main__\":\n    recorder = AudioToTextRecorder()\n\n    while True:\n        recorder.text(process_text)\n```\n\n## External Audio\n\nSet `use_microphone=False` when audio comes from a file, stream, websocket, or\nanother process. Feed 16-bit mono PCM chunks at 16 kHz, or pass the original\nsample rate so RealtimeSTT can resample:\n\n```python\nfrom RealtimeSTT import AudioToTextRecorder\n\nif __name__ == \"__main__\":\n    recorder = AudioToTextRecorder(use_microphone=False)\n\n    with open(\"audio_chunk.pcm\", \"rb\") as audio_file:\n        recorder.feed_audio(audio_file.read(), original_sample_rate=16000)\n\n    print(recorder.text())\n    recorder.shutdown()\n```\n\nMore examples are in [docs\u002Fquick-start.md](docs\u002Fquick-start.md) and\n[docs\u002Fexternal-audio.md](docs\u002Fexternal-audio.md).\n\n## Configuration Reference\n\nEvery `AudioToTextRecorder` constructor parameter is documented in\n[docs\u002Fconfiguration.md](docs\u002Fconfiguration.md), including model\u002Fengine\nselection, realtime transcription, VAD timing, wake words, callbacks, external\naudio, logging, and executor injection.\n\n## Features\n\n- Voice activity detection with WebRTC VAD and Silero VAD.\n- Final and realtime transcription with selectable engines.\n- Optional wake word activation through Porcupine or OpenWakeWord.\n- Direct microphone input or application-fed audio chunks.\n- Event callbacks for recording, VAD, realtime text, transcription, and wake\n  word state.\n- A FastAPI browser streaming server example with multi-user session isolation,\n  shared inference resources, metrics, and health endpoints.\n\n## Documentation\n\n- [Quick start](docs\u002Fquick-start.md): shortest demos and common recording\n  patterns.\n- [Installation](docs\u002Finstallation.md): platform setup, CUDA notes, and optional\n  dependencies.\n- [Configuration](docs\u002Fconfiguration.md): complete `AudioToTextRecorder`\n  parameter reference.\n- [Transcription engines](docs\u002Ftranscription-engines.md): engine selection and\n  setup links.\n- [Wake words](docs\u002Fwake-words.md): Porcupine and OpenWakeWord setup.\n- [External audio](docs\u002Fexternal-audio.md): feeding audio without a microphone.\n- [Testing](docs\u002Ftesting.md): maintained unit and opt-in golden test workflow.\n- [Test scripts](docs\u002Ftest-scripts.md): demos, manual tests, regressions, and\n  legacy experiments under `tests\u002F`.\n- [FastAPI server](docs\u002Ffastapi-server.md): browser server configuration,\n  protocol, metrics, and deployment notes.\n- [Troubleshooting](docs\u002Ftroubleshooting.md): common install, audio, CUDA,\n  model, dependency, and runtime errors.\n\nEngine-specific references:\n\n- [faster-whisper](docs\u002Fengines\u002Ffaster-whisper.md)\n- [whisper.cpp](docs\u002Fengines\u002Fwhisper-cpp.md)\n- [OpenAI Whisper](docs\u002Fengines\u002Fopenai-whisper.md)\n- [Moonshine](docs\u002Fengines\u002Fmoonshine.md)\n- [sherpa-onnx](docs\u002Fengines\u002Fsherpa-onnx.md)\n- [Parakeet NeMo](docs\u002Fengines\u002Fparakeet-nemo.md)\n- [Transformers engines](docs\u002Fengines\u002Fhf-transformers.md)\n- [Cohere Transcribe](docs\u002Fengines\u002Fcohere.md)\n\n## Server Example\n\nThe browser server lives in `example_fastapi_server`:\n\n```bash\npython -m pip install -r example_fastapi_server\u002Frequirements.txt\npython example_fastapi_server\u002Fserver.py --host 0.0.0.0 --port 8010\n```\n\nOpen `http:\u002F\u002Flocalhost:8010`. See [docs\u002Ffastapi-server.md](docs\u002Ffastapi-server.md)\nfor engine recipes, websocket protocol details, health checks, and metrics.\n\n## Contributing\n\nFocused tests and small changes are easiest to review. The project keeps fast\nunit tests separate from opt-in real-model tests; see [docs\u002Ftesting.md](docs\u002Ftesting.md).\n\n## License\n\nMIT\n\n## Author\n\nKolja Beigel\n","RealtimeSTT 是一个用于实时语音转文字的Python库，它具备先进的语音活动检测、唤醒词激活以及即时转录功能。项目的核心特性包括使用WebRTC VAD和Silero VAD进行精准的语音活动检测、支持多种可选引擎实现最终及实时转录，并且能够通过Porcupine或OpenWakeWord设置唤醒词。此外，该库还允许直接从麦克风输入或应用程序提供的音频片段中读取数据，并提供了丰富的事件回调选项以满足不同需求。RealtimeSTT适用于需要快速准确地将语音转换为文本的应用场景，如智能助手、听写工具、浏览器流媒体服务器等。其简洁易用的设计使得开发者仅需几行代码即可集成强大的语音处理能力到自己的项目中。",2,"2026-06-11 03:35:24","high_star"]