vllm-omni/docs/contributing/model

Adding a New Model

This section provides comprehensive guidance on how to add a new model to vLLM-Omni.

Documentation

• Adding a New Omni Model Guide: Complete step-by-step guide using Qwen3-Omni as an example

The guide covers:

• Directory structure and organization

• Implementing stage components (thinker, talker, code2wav)

• Creating the unified model class

• Model registration

• Stage configuration

• Stage input processors

• Testing strategies

• Adding a New Diffusion Model Guide: Complete step-by-step guide using Qwen/Qwen-Image-Edit as an example The guide covers:

• Overview

• Directory Structure

• Step-by-Step Implementation

• Testing

Quick Start

For a quick reference, see the Adding a New Multi-Stage Model Guide and Adding a New Diffusion Model Guide.