Description
Vidu S1 API is a commercial-grade streaming video generation model designed for real-time interactive digital humans. This innovative API allows developers to create live sessions where an AI character can see, hear, and converse with users. The platform supports unlimited-duration generation, offering over 50 voices in 28 languages, making it versatile for various applications.
The billing for Vidu S1 API begins as soon as the digital character is ready and the session enters the 'on_live' state. The usage is metered at a rate of 3 credits per 2 seconds, deducted every 6 seconds, ensuring that users only pay for the time they actively use the service. Each session can last up to 600 seconds, and for longer interactions, users can create new sessions without quality loss.
Vidu S1 is not limited to HTTP interactions; it utilizes a combination of HTTP for session management, AliRTC for real-time audio and video transport, and WebSocket for control signaling. This integration allows for a seamless user experience where the digital character can interact in quasi real-time, responding to user inputs with expression and personality.
The API is designed for various industries, including AI companionship, virtual idols, training and education, customer service, live-stream commerce, and interactive entertainment. It enables teams to deploy production-grade digital humans rapidly, transforming user engagement across platforms. With features like persona memory and multimodal perception, Vidu S1 creates personalized and consistent interactions, enhancing user satisfaction.
In addition to its robust capabilities, Vidu S1 API offers a free trial with 1,000 credits for new users, allowing them to explore the platform's features without initial investment. This makes it an attractive option for developers looking to integrate advanced AI interactions into their products quickly.
Vidu S1 API's Core Features
Unlimited Interactive Duration
50+ Voices
28 Languages Supported
Real-Time Interaction
Session Management via HTTP
Audio/Video Transport via AliRTC
Control Signaling via WebSocket
Custom Voice Cloning
Short-Term Memory for Personas
High-Resolution Output
How to use Vidu S1 API?
Create a Session: Make a POST call with the character's persona, avatar image, and voice to receive a session ID.
Join the RTC Channel: Use the returned token to join the AliRTC channel and publish the user's microphone and camera.
Open the WebSocket: Connect to the control channel, passing the authentication token in the query string.
Wait Until Ready: Monitor the connection for a success acknowledgment indicating the character is live.
Keep the Session Alive: Respond to server pings and handle any disconnect messages appropriately.
Hang Up & Query Billing: Send a hangup message, close the WebSocket, and query the final session status.
Vidu S1 API's Use Cases
- AI Companionship
- Virtual Idols
- Training & Education
- AI Customer Service
- Live-Stream Commerce
- Interactive Entertainment
