{"name":"RealtimeVAD","repo":"KoljaB/RealtimeVAD","url":"https://midorreal.com/project/koljab-realtimevad","repository":"https://github.com/KoljaB/RealtimeVAD","description":"Faster speech off signals for spec gen voice pipelines","facts":{"license":{"value":"MIT","source":"https://github.com/KoljaB/RealtimeVAD","checked":"20 Nov 2026"},"language":{"value":"Python","source":"https://github.com/KoljaB/RealtimeVAD","checked":"20 Nov 2026"},"last_release":{"value":"v0.1.1, 14 Nov 2026","source":"https://github.com/KoljaB/RealtimeVAD/releases","checked":"20 Nov 2026"},"releases":{"value":"4","source":"https://github.com/KoljaB/RealtimeVAD/releases","checked":"19 Nov 2026"},"contributors":{"value":"1","source":"https://github.com/KoljaB/RealtimeVAD/graphs/contributors","checked":"19 Nov 2026"},"archived":{"value":"no","source":"https://github.com/KoljaB/RealtimeVAD","checked":"20 Nov 2026"}},"stars":99.0,"writeup":{"what_it_is":"RealtimeVAD is a Python voice activity detector that keeps Silero VAD for speech detection and adds an earlier \"pause\" signal based on audio level compared with room noise and the speaker's level. It emits start, pause, resume, end and onset events, and runs with numpy and onnxruntime.","audience":"Developers building voice assistants and speech pipelines who need early speech-off detection","claims":[{"kind":"specific","claim":"Gives a pause signal a median 92 ms before Silero's probability drops","excerpt":"`pause` comes a median 92 ms before Silero's probability drops","status":"not_checked"},{"kind":"specific","claim":"Fast mode leads by 130 ms on the measured turns","excerpt":"`pause_mode=\"fast\"` leads by 130 ms on the same turns","status":"not_checked"},{"kind":"specific","claim":"Uses about 8 ms of CPU per second of audio on one thread","excerpt":"one CPU thread, about 8 ms per second of audio","status":"not_checked"},{"kind":"specific","claim":"Needs only numpy and onnxruntime, with the Silero model bundled and no torch","excerpt":"Needs only numpy and onnxruntime; the Silero model is bundled, no torch.","status":"not_checked"},{"kind":"specific","claim":"Provides a drop-in VADIterator with Silero's interface","excerpt":"a `VADIterator` with Silero's interface","status":"not_checked"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-11-14"},"why_now":null,"owner_supplied":null,"score":{"verdict":"overhyped","hype":88,"reality":51,"gap":37,"momentum":null,"confidence":85,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T07:00:00+00:00","verdict_source":"rule"},"early_read":null,"changes":[{"date":"2026-11-14T21:38:25+00:00","text":"Latest release v0.1.1","source":"https://github.com/KoljaB/RealtimeVAD/releases"},{"date":"2026-11-12T00:00:25+00:00","text":"First release","source":"https://github.com/KoljaB/RealtimeVAD/releases"},{"date":"2026-11-07T15:45:35+00:00","text":"Repository created","source":"https://github.com/KoljaB/RealtimeVAD"}],"cite":"RealtimeVAD is a GitHub project at https://github.com/KoljaB/RealtimeVAD written mainly in Python, described there as \"Faster speech off signals for spec gen voice pipelines\".","attribution":"Data from Mid or Real, https://midorreal.com/project/koljab-realtimevad, checked 22 November 2026. Attribution with a link is required.","still_real":null}