Superinterface logo
Superinterface
ProductPricingExamplesDocsBlog
GitHub
Discord
Book a demo
Getting started
AI Provider Setup
OpenAI
Anthropic
Azure OpenAI
FAQ
Assistants
Create assistant
Functions
API Request
Client tool
Firecrawl
Multi-agents
Interfaces
Create interface
Interaction Modes
Realtime Voice Chat
Text-to-Speech Voice Chat
AI Builder
Publish interface
Components
Annotations
Concepts
Why Superinterface
How it works
API reference
React
Installation
Components
Hooks
JavaScript

How to Enable Voice Assistant with Realtime API for Your Assistant

Want to create a truly seamless conversational experience? Superinterface now supports OpenAI’s Realtime API, allowing for incredibly fast, natural voice interactions with your AI assistant.
Unlike traditional text-based exchanges, this API uses WebRTC to establish a direct peer-to-peer connection, delivering ultra-low latency responses.
This means your assistant can listen, process, and reply in real-time—creating a smooth, uninterrupted dialogue for your users.
Let’s get started with setting it up step by step.

Step 1: Setting Up Your Assistant

Create and Configure

    Create a New Assistant
    Log in to your Superinterface dashboard, navigate to Assistants, and click New Assistant.
    For detailed instructions, check out how to create an assistant.
    Choose OpenAI as Your AI Provider
    In the setup screen, select OpenAI as your AI provider.
    Set Thread Storage and Execution to Superinterface Cloud
    This step is essential since OpenAI Assistants API doesn’t yet support Realtime models. Superinterface Cloud bridges this gap by managing thread storage and execution for you.
Setting up a voice assistant with OpenAI's Realtime API in Superinterface.
    Select a Realtime Model
    From the model dropdown, choose a model with “realtime” in its name. Examples include:
    GPT-4.0 Mini Realtime Preview
    GPT-4.0 Realtime Preview
    Versions with locked dates, like GPT-4.0 Mini Realtime Preview 2024-12-17.

Add Personalization

    Customize Your Assistant
    Name your assistant and define its behavior with detailed instructions.
    Set an Initial Message
    In the Initial Messages section, create a welcoming first interaction (e.g., “Hi! How can I assist you?”). This message will be spoken aloud when the voice chat starts.
    Save Your Assistant
    Once everything is configured, click Save to finalize your setup.

Step 2: Publish Your Realtime Voice Interface

Creating an interface to your realtime voice assistant on Superinterface.
Creating an interface to your realtime voice assistant on Superinterface.
Once your assistant is set up, it’s time to enable Realtime voice chat and publish the interface for users to access.

Enable Voice Chat

    Create a New Interface
    Navigate to the Publish tab and select Choose Interface. Click Create New Interface to get started.
    Select Interaction Mode: Voice Chat
    In the interaction mode selection, choose Voice Chat.
    A confirmation will appear, letting you know that OpenAI Realtime API and WebRTC are enabled for this setup.
    Enabling Voice Mode in Superinterface.
    Enabling Voice Mode in Superinterface.

Publish Your Assistant

    Finalize Publishing
    Choose your preferred publishing method (e.g., subdomain, custom domain, script tag, or React components) and click Publish.
    For step-by-step guidance, visit how to publish an interface.
When users access your assistant, a WebRTC peer-to-peer connection directly links them to OpenAI’s real-time API. This ensures the fastest, most seamless experience possible.
Voice assistant powered by OpenAI's Realtime API via Superinterface.
Voice assistant powered by OpenAI's Realtime API via Superinterface.

You’re all set!

Now go ahead and let your users experience seamless, real-time voice interactions with your intelligent AI assistant.