This guide provides comprehensive instructions for implementing voice-enabled chat functionality on websites. The feature allows visitors to engage in real-time voice conversations with AI agents directly through their browsers.

Key Setup Steps:

The process involves four primary phases: enabling the feature through Labs settings, configuring widget parameters with a custom agent name, obtaining and installing embed code, and testing visitor interactions.

Core Functionality:

The widget enables real-time voice conversations directly inside the chat widget without requiring phone numbers or app downloads. Users can start sessions by clicking a microphone icon, mute/unmute as needed, and disconnect at will.

Notable Features:

• Support for up to 100 Voice AI agents
• Built-in reCAPTCHA protection against repeated misuse
• Browser-based microphone and speaker access
• Non-recording transparency notices
• Compatibility with existing Voice AI workflows (except call transfer)

Important Requirement:

Users cannot create a voice AI chat widget unless your Voice AI setup along with agent creation is completed beforehand.

The guide includes troubleshooting sections, best practices for agent naming and page placement, and a practical lead-capture implementation example.