Introduction
Welcome to the Lemon Slice API reference. Our API enables seamless integration of Lemon Slice’s state-of-the-art, hyper-expressive talking-head AI videos into your applications.
Lemon Slice offers two primary API frameworks: Video Generation and Video Streaming.
Video Generation
The Video Generation API provides RESTful endpoints for animating and lip-syncing images or videos. Clients initiate a video generation request by submitting the necessary payload and then poll the corresponding status endpoint to monitor progress. Generation time varies based on input complexity and may range from a few seconds to several minutes.
Video Streaming
The Video Streaming API is designed for real-time, interactive applications that require expressive lip-sync. Clients establish a session for a specific character and stream audio data over a WebSocket connection. A Daily.co URL is provided for seamless integration of the streaming video into supported frameworks. Note: the Video Streaming API is currently an audio-in, video-out interface. It does not handle speech-to-text or LLM response generation. It is up to the developer to determine the audio that should be used in the generated video stream.
Refer to each endpoint’s documentation for details on usage and pricing.