OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Audio & Voice AI

Cobra

Cobra is Picovoice’s deep-learning voice activity detection engine for detecting speech activity in audio on-device. Its repository provides SDKs and demos for incorporating the engine into audio applications.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Cobra — OpenViST project profile in Audio & Voice AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Cobra when an audio application needs a voice-activity signal to distinguish speech-active audio from silence, and the developer can supply audio frames through one of the documented SDKs.

Who it is for

  • Developers adding speech-activity detection to recording or audio-processing applications.
  • Teams evaluating voice detection in a supported SDK or demo environment.

Documented uses

  • Identify speech-active audio frames in a recording or incoming audio stream.
  • Use voice activity as a signal for separating speech from silence in an audio workflow.
  • Try real-time microphone-based voice detection through a documented demo.
Before you choose

Requirements & limits

  • The Python SDK requires an AccessKey from Picovoice Console.
  • The Python SDK exposes the required sample rate and frame length through the initialized handle.
  • The repository lists SDKs and demos for Python, .NET, C, Android, iOS, web and Node.js.
  • The repository identifies the project license as Apache-2.0.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Try real-time voice detection with the Python microphone demo

Run the documented Python demo to process microphone audio and report detected voice activity.

Before you start

  • Python package installation using pip3
  • A working microphone connected to the device
  • An AccessKey obtained from Picovoice Console

Steps

  1. Install the demo package with sudo pip3 install pvcobrademo.

  2. Run cobra_demo_mic --access_key ${AccessKey}, replacing the placeholder with the AccessKey from Picovoice Console.

  3. Speak into the microphone and observe the terminal output for detected voice activity.

Expected outcome

The demo processes microphone input in real time and reports voice activity in the terminal.

Guide limitations

  • This is a microphone demo, not a documented recipe for integrating Cobra into an application.
  • The instructions require a working microphone and a Picovoice AccessKey.
Documentation 1 ↗
Source-based practical guide

Process audio frames with the Python SDK

Initialize the Python engine and pass audio frames to obtain voice-probability values.

Before you start

  • The pvcobra Python SDK installed with pip3
  • An AccessKey obtained from Picovoice Console
  • Audio frames supplied by the application

Steps

  1. Install the SDK with the command below.

    Documented command
    pip3 install pvcobra

  2. Import pvcobra and create the engine with pvcobra.create(access_key=${AccessKey}).

  3. Read the required sample rate from handle.sample_rate and frame length from handle.frame_length.

  4. Pass each audio frame to handle.process(...) and use the returned voice-probability value.

  5. Release the engine resources with handle.delete() when finished.

Expected outcome

The application can obtain a voice-probability value for each processed audio frame.

Guide limitations

  • The SDK example leaves audio-frame acquisition to the application.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
Picovoice/cobra · Official homepage title · retained in its source language
How we review sourcesSuggest a correction