ojin-client 0.6.5

Ojin platform services

Ojin Speech-To-Video Client

A WebSocket client for Ojin STV model that handles communication with Inference Server. It is used for example by Ojin Video Service for Pipecat https://github.com/pipecat-ai/pipecat

Requirements

• Python 3.10+
• pip or uv package manager

Installation

pip install ojin-client

Usage

from ojin.ojin_client import OjinClient

avatar = OjinClient(
        ws_url="THE_OJIN_URL",
        api_key="YOUR_OJIN_API_KEY",
        avatar_config_id="YOUR_OJIN_CONFIGURATION"M
    )

API methods

• connect: Establishes the websocket connection with the ojin inference server
• start_interaction: Creates a UUID to send in the interaction messages
• send_message: Sends messages to the ojin inference server
• receive_message: Receives messages from the ojin inference server
• close: closes the connection with ojin inference server

Messages

• OjinInteractionReadyMessage: Message indicating that an interaction is ready to begin

• OjinSessionReadyMessage: Message to start a new interaction with an Ojin Persona

• OjinCancelInteractionMessage: Message to cancel an interaction

• OjinAudioInputMessage: Contains the audio data as bytes and a flag indicating if this is the last input

Responses

• StartInteractionResponseMessage: Response message for a successful interaction start

• OjinInteractionResponseMessage: Response message containing video data from the avatar

• ErrorResponseMessage: Response message informing the client there was an error