# TEN-framework/ten-vad

Voice Activity Detector (VAD) : low-latency, high-performance and lightweight

Repository: https://github.com/TEN-framework/ten-vad
Canonical: https://ross.abutalabs.com/products/ten-vad
Homepage: https://huggingface.co/TEN-framework/ten-vad
Language: C
License: NOASSERTION
License Family: other
Topics: conversational-ai, real-time, speech-processing, vad, voice-activity-detection, voice-commands, voice-recognition, audio, automatic-speech-recognition, speech, silero-vad, voice-agent
Last push: 2026-02-02T02:51:42+00:00

## Health v2 (maintenance only)
Score: 50/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 65, release rhythm 40, longevity 35
- inputs: {"age_days": 491, "days_push": 212, "days_rel": 418, "gap_med": 0, "n_releases_24m": 2}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2248, forks 177 (observed 2026-08-28T04:06:30.269399+00:00)

## What it is
TEN VAD is a lightweight, low-latency voice activity detection library with an ONNX model for real-time speech detection. It supports C, Python, Go, Java, and JavaScript/WASM across desktop, mobile, and browser platforms.

## Use cases
- detect speech segments in a real-time audio stream
- build a voice agent that knows when the user stops talking
- replace silero-vad with a faster lower-latency VAD
- trim silence from audio before running ASR
- run voice activity detection in the browser with wasm
- gate wake-word or voice-command pipelines with speech detection

## When to choose
- you need low-latency, real-time VAD for conversational AI or voice agents
- you need cross-platform deployment including mobile and browser via ONNX/WASM
- you want a lightweight model with bindings for C, Python, Go, Java, or JS

## When to avoid
- you need full speech recognition or transcription rather than just speech detection
- you require a permissively licensed library, since the license is non-standard
- you need offline batch audio analysis with rich tooling rather than streaming inference

## Facets
- artifact type: library
- maturity: active
- function: audio-processing, speech-recognition, machine-learning
- domain: speech-processing, artificial-intelligence
- platform: windows, browser, wasm, python, cpp, cross-platform
- tags: vad, voice-activity-detection, onnx, conversational-ai, voice-agent, silero-vad-alternative, low-latency, real-time, audio, linux, macos, android, ios

## Member repositories
- TEN-framework/ten-vad (main) score 50

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:30.269399+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:43:55.674381+00:00, confidence not recorded.
  - readme: https://github.com/TEN-framework/ten-vad (fetched 2026-08-28T04:06:30.269399+00:00, sha 7887faade828)
  - homepage: https://huggingface.co/TEN-framework/ten-vad (fetched 2026-08-29T10:24:08.758601+00:00, sha 1f8e634f136b)
  - site_page: https://huggingface.co/docs (fetched 2026-08-29T10:24:08.762613+00:00, sha bdec26667b98)
  - site_page: https://huggingface.co/docs/inference-providers (fetched 2026-08-29T10:24:08.766987+00:00, sha 8a5d0f819473)
  - site_page: https://huggingface.co/docs/hub/model-cards (fetched 2026-08-29T10:24:08.769273+00:00, sha 60ded09a56b0)
  - site_page: https://huggingface.co/pricing (fetched 2026-08-29T10:24:08.764464+00:00, sha de6b7a178be5)
  - site_page: https://huggingface.co/huggingface (fetched 2026-08-29T10:24:08.771372+00:00, sha 0a25e6fef275)
- Data as of 2026-08-30T08:39:29.467469+00:00.