NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI
PyPI, #2826 most downloaded on PyPI
Voice Activity Detector (VAD) by Silero
Latest stable
6.2.3
Last release 2 months ago
23 Sep 2026
Ships fairly regularly
a new release about every 4 months
Nearly every release is documented
notes for 8 of 8 stable releases
Nothing withdrawn
no release was ever pulled
2 years old
10 releases · first in 2024
Drag across the ruler to compare, or tap a release to open it.
torchaudio is now an optional dependency instead of required. Only torch is required — the model itself never used torchaudio, only the read_audio / s
torchaudio is now an optional dependency instead of required. Only torch is required — the model itself never used torchaudio, only the read_audio / save_audio helpers did.
Users who need audio I/O must now install a backend explicitly:
pip install silero-vad[audio] (torchaudio)pip install silero-vad[codec] (torchcodec)pip install silero-vad[all] (everything)Key changes:
read_audio / save_audio now work with either torchaudio or torchcodec; with neither installed they raise an error naming the extra to installimport silero_vad failed when onnxruntime was not installed — the sequence path is now imported lazilyIf you load audio yourself, no backend is needed at all:
from silero_vad import load_silero_vad, get_speech_timestamps
model = load_silero_vad()
audio = ... # 1-D float32 torch.Tensor from any source
ts = get_speech_timestamps(audio, model, return_seconds=True)Your coding agent can read these notes before it upgrades. Set up the MCP server →
Added offline ONNX sequence to release GIL- #795
Added offline ONNX sequence to release GIL- #795
silero_vad_16k_sequence.onnx modelfrom silero_vad import load_silero_vad, get_speech_timestamps_sequence, read_audio
model = load_silero_vad(sequence=True) # loads silero_vad_16k_sequence.onnx
audio = read_audio("audio.wav", sampling_rate=16000)
ts = get_speech_timestamps_sequence(audio, model, return_seconds=True)onnxruntime is now an optional dependency instead of required.
You now need to install ONNX Runtime explicitly:
pip install onnxruntime # for CPU
pip install onnxruntime-gpu # for GPU
VAD training paradigm reworked;
Full Changelog: https://github.com/snakers4/silero-vad/compare/v6.1...v6.2
Currently we provide the following endpoints:
model= Params Model type Streaming Languages PyTorch ONNX opset Colab
'silero_vad' ~260K VAD Yes 6000+ ✔️ 15 ✔️, 16 ✔️
model= Params Model type Streaming Languages PyTorch ONNX Colab
'silero_number_detector' 1.1M Number Detector No ru , en , de , es ✔️ ✔️
'silero_lang_detector' 1.1M Language Classifier No ru , en , de , es ✔️ ✔️
'silero_lang_detector_95' 4.7M Language Classifier No 95 languages ✔️ ✔️
You can’t perform that action at this time.
Remove ipdb and raise error directly in get_speech_timestamps by @mpariente in https://github.com/snakers4/silero-vad/pull/695
Full Changelog: https://github.com/snakers4/silero-vad/compare/v6.0...v6.1
v5 features and improvements kept;
Full Changelog: https://github.com/snakers4/silero-vad/compare/v5.1.2...v6.0
A tag to upload new PIP package.
A tag to upload new PIP package.
Experimental pip-package release;
Full Changelog: https://github.com/snakers4/silero-vad/compare/v5.0...v5.1
Nothing published for this version
Nothing published for this version
One column per month.