skillsafe-ai/silero-vad

Silero VAD v5

Browser-ready import artifacts for voice-activity-detection, produced by SkillSafe's reproducible converter (models/ in skillsafe.ai) from a pinned upstream source. Every byte here is derivable from that source plus the recipe below; nothing was edited by hand.

Provenance

Upstream	https://raw.githubusercontent.com/snakers4/silero-vad/60b7ffa243625ebdc1070275a29f18c87843786a/src/silero_vad/data/silero_vad.onnx
Upstream SHA-256 / commit	60b7ffa243625ebdc1070275a29f18c87843786a
Recipe	recipes/silero-vad.yaml — sha256 add4c06a8242f6db1760bf32ab2de236c33aedb774322927e2f57108ca41e652
Toolchain	Python 3.12.13, torch 2.10.0, onnx 1.23.0, onnxruntime 1.30.0 on Darwin 25.6.0 arm64
Converted	2026-09-22T18:56:06+00:00

Files

file	class	size	SHA-256
silero_vad.onnx	registry	2.22 MB	1a153a22f4509e292a94e67d6f9b85e8deb25b4988682b7e174c65279d8788e3
silero_vad_half.onnx	registry	1.22 MB	1e0b195ad4806595ef4466f419d16fca7e4afcfc6669b8c0b5f76ea87547c769

registry files are parameter files served from models.skillsafe.ai once vetted; bundle files ship inside an app; registry-shared is a runtime library reused by every model of the same architecture.

Verification

Imported as published upstream (no conversion). Each file is pinned by SHA-256 to its source; every ONNX file passed onnx.checker and a CPU smoke run under onnxruntime with zero-filled inputs at the declared shapes:

file	inputs	outputs	ms
silero_vad.onnx	input[1, 576], state[2, 1, 128], sr[]	output[1, 1], stateN[2, 1, 128]	0.3
silero_vad_half.onnx	input[1, 576], state[2, 1, 128]	output[1, 1], stateN[2, 1, 128]	0.2

Use in the browser

import * as ort from "onnxruntime-web";
const session = await ort.InferenceSession.create("https://huggingface.co/skillsafe-ai/silero-vad/resolve/main/silero_vad.onnx", { executionProviders: ["webgpu", "wasm"] });

Contract (silero_vad.onnx): input input float32 ['?', '?'], state float32 [2, '?', 128], sr int64 [] → output output float32 ['?', 1], stateN float32 ['?', '?', '?']. Opset 16.

Licence and attribution

Silero VAD: Copyright (c) 2020-present Silero Team. MIT License. https://github.com/snakers4/silero-vad

Licence: MIT — notice: https://github.com/snakers4/silero-vad/blob/60b7ffa243625ebdc1070275a29f18c87843786a/LICENSE. The conversion recipe and this model card are part of the SkillSafe repository and carry its licence; the weights remain under the upstream licence above.

The full manifest.json in this repo records the recipe, sources, toolchain (including the uv.lock hash) and per-file verification numbers.