Realtime Voice

What is Realtime Voice?

Realtime Voice is Genspark's voice-first AI assistant that lets you talk directly to Super Agent using natural speech. Powered by real-time bidirectional audio, it enables hands-free interaction for research, content creation, task delegation, and more — all through conversation.

• Talk, don't type — Speak naturally to Super Agent and hear instant voice responses, just like a phone call
• Multitask with voice — Request complex tasks (slides, docs, images, deep research) while continuing the conversation — tasks run in the background without interrupting your flow
• Task tracking — View and manage all background tasks from a dedicated panel, with real-time progress and one-tap access to results
• Seamless mode switching — Switch between voice and text input anytime during a session

Quick Start

Get started in three simple steps:

1. Start a Super Agent session

   Go to the Genspark homepage and start a new Super Agent session. In the chat input box, you'll see a voice button (black circle icon) on the right side — click it to enter Realtime Voice mode.

2. Grant microphone access

   Allow browser microphone permission when prompted. The connection is established automatically.

3. Start talking

   Speak your request naturally. The AI responds in real time with both voice and text. You can also switch to text input at any time using the keyboard button.

Feature Guide

Background Tasks

One of Realtime Voice's most powerful features is the ability to launch background tasks while continuing your conversation. Simply describe what you need, and Super Agent will automatically delegate the work to a specialized agent.

Supported task types include:

• Slides — "Create a presentation about..."
• Docs — "Write a report on..."
• Images — "Generate an image of..."
• Deep Research — "Do deep research on..."
• Websites — "Build a landing page for..."
• Spreadsheets — "Create a spreadsheet analyzing..."

You can launch multiple tasks at once — each runs independently in the background.

Background Tasks Panel

During a voice session, a task button appears in the voice control bar at the bottom of the screen. When tasks are running, a badge on the button shows the number of active tasks.

Click the task button to open the Background Tasks panel:

• On desktop, it slides in from the right side of the screen
• On mobile, it appears as a bottom sheet

The panel shows all your background tasks with their current status — running, completed, or failed. Click any task to jump directly to its results.

Voice & Text Switching

You can switch between voice and text input at any time during a session:

• Voice to text — Click the keyboard icon to type your message instead
• Text to voice — Click the microphone icon to switch back to voice mode

Your conversation history is preserved regardless of input mode.

FAQ

Q: Does Realtime Voice cost credits?

Yes, Realtime Voice sessions consume credits based on usage duration. Check the Credits Guide for detailed pricing.

Q: Can I use Realtime Voice on mobile?

Yes, Realtime Voice works on mobile browsers. Tap the voice button in the chat input box to start a session.