pymicro-vad
Self-contained voice activity detector
Decision gist · record as of 2026-08-14
Yes, if you need a lightweight, offline voice activity detector for real-time audio processing. The package is actively maintained, has no known vulnerabilities, uses a permissive license, and provides pre-built wheels for common platforms. The main constraint is the strict audio format requirement (10ms chunks of 16-bit mono PCM at 16kHz), which you must handle yourself.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.9+; audio input must be exactly 160 samples (10ms at 16kHz) of 16-bit mono PCM data per call.
- Medium install friction due to compiled wheels; the package provides pre-built binaries for x86_64 and aarch64 on manylinux platforms.
- Actively maintained with a recent release.
License · maintenance · safety
Apache (permissive) — Licensed under Apache (permissive), so you can use it freely in commercial and open-source projects without copyleft obligations.
last release 2026-06-29 (46 days) · last repo commit 2026-06-29 · 37 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 119,537 downloads/mo, #12,068 on PyPI
Alternatives
Verify before relying
pip install pymicro-vad
from pymicro_vad import MicroVad
vad = MicroVad()
threshold = 0.5
# Process 10ms chunks of 16-bit mono PCM @16kHz
while audio := get_10ms_of_audio():
speech_prob = vad.process_10ms(audio)
if speech_prob > threshold:
print("Speech detected")- Whether the model weights are included in the wheel or downloaded separately on first use.
- Latency and CPU overhead of processing 10ms chunks in real-time scenarios.
- Accuracy of speech detection across different acoustic environments and speaker profiles.
What it is and what it does
pymicro-vad is a lightweight, self-contained voice activity detector built on machine learning architecture from microWakeWord. It takes small chunks of audio (10ms at a time) and returns a probability score indicating whether that chunk contains speech or silence. The detector is designed to run locally without external dependencies or network calls, making it suitable for embedded systems, edge devices, or applications where low latency and offline operation are important.
You feed it raw 16-bit mono PCM audio at 16kHz sample rate, 160 samples per call (exactly 10ms), and it returns a speech probability. Scores below zero indicate the detector needs more audio to make a decision, scores above your chosen threshold indicate speech, and scores in between indicate silence. The package ships with pre-compiled wheels for common Linux platforms, though building from source requires python3-dev and build-essential.
Use it for
- Real-time speech detection in voice assistant applications to trigger wake-word detection only when speech is present.
- Audio preprocessing in transcription pipelines to skip silence and reduce processing of non-speech segments.
- Endpoint detection in voice recording applications to automatically stop recording when silence is detected.
- Edge device deployment where a lightweight, offline VAD is needed without cloud dependencies.
- Audio stream filtering in IoT devices or embedded systems with limited computational resources.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you need a lightweight, offline voice activity detector for real-time audio processing.
The package is actively maintained, has no known vulnerabilities, uses a permissive license, and provides pre-built wheels for common platforms. The main constraint is the strict audio format requirement (10ms chunks of 16-bit mono PCM at 16kHz), which you must handle yourself.
Install
pymicro-vad on PyPI
Before you install
Medium install friction due to compiled wheels; the package provides pre-built binaries for x86_64 and aarch64 on manylinux platforms. Actively maintained with a recent release.
Requires Python 3.9+; audio input must be exactly 160 samples (10ms at 16kHz) of 16-bit mono PCM data per call.
License in practice
Licensed under Apache (permissive), so you can use it freely in commercial and open-source projects without copyleft obligations.
Quickstart
pip install pymicro-vad
from pymicro_vad import MicroVad
vad = MicroVad()
threshold = 0.5
# Process 10ms chunks of 16-bit mono PCM @16kHz
while audio := get_10ms_of_audio():
speech_prob = vad.process_10ms(audio)
if speech_prob > threshold:
print("Speech detected")Verify before relying
- Whether the model weights are included in the wheel or downloaded separately on first use.
- Latency and CPU overhead of processing 10ms chunks in real-time scenarios.
- Accuracy of speech detection across different acoustic environments and speaker profiles.
Package facts
| License | Apache permissive |
| Python support | Supports the current Python release >=3.9 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 46 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 119,537 / month, #12,068 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: pymicro_vad-2.1.0-cp39-abi3-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; pymicro_vad-2.1.0-cp39-abi3-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “real-time speech detection”
- pymicro-vadA self-contained voice activity detector that processes 10ms chunks…
- realtimesttRealtimeSTT converts speech from microphone or external audio streams…
- speechmatics-voicePython SDK for building real-time voice applications with automatic…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also webrtcvad · webrtcvad-wheels · pyrnnoise · silero-vad · pvporcupine · realtimestt · openwakeword · aic-sdk · whisper-timestamped · livekit-plugins-turn-detector