Skip to content

Summary

Entry Points

Main entry points for vLLM-Omni inference and serving.

Clients

Client libraries for full-duplex sessions: the Realtime WebSocket client, the in-process client over DuplexOmni, and the per-model session presets.

Inputs

Input data structures for multi-modal inputs.

Engine

Engine classes for offline and online inference.

Core

Core scheduling and caching components.

Configuration

Configuration classes.

Workers

Worker classes and model runners for distributed inference.