docs/index.md

# Welcome to vLLM-omni

**vLLM-omni** extends vLLM capabilities to support multi-modality model inference and serving, particularly focusing on non-autoregressive structures and non-textual outputs. It aims to support:

- **Multi-modal Models**: Text, image, video, audio, and sensor data processing
- **Non-autoregressive Architectures**: Diffusion Transformers (DiT) and other parallel generation models
- **Heterogeneous Outputs**: Beyond traditional text generation to multimodal outputs

## <span class="twemoji">🎯</span> Key Features

### Multi-Engine Support

- **Autoregressive Engine**: Traditional text generation with enhanced KV-caching
- **Diffusion Engine**: Support for DiT models and iterative generation
- **Hybrid Engine**: Combined AR+DiT processing pipelines

### Modality Processing

- **Text**: Advanced tokenization and embedding generation
- **Image**: Vision encoder integration (CLIP, etc.)
- **Audio**: Speech processing and audio embedding

## <span class="twemoji">🤖</span> Supported Models

- Qwen-omni (Thinker-Talker-Codec structure)
- HunyunaImage 3.0 (Ongoing)
- Qwen-Image (Ongoing)

## <span class="twemoji">📚</span> Documentation Navigation

- To run open-source models on vLLM-Omni, we recommend starting with the [:material-code-tags: User Quide](user_guide/getting_started/quickstart.md)
- To develop and contribute to vLLM-Omni, we recommend starting with the [:material-tools: Developer Guide](contributing/README.md)
[Docs] Setup Documentation System and Re-organize Dependencies (#49 ) 2025-11-15 21:45:43 +08:00			`# Welcome to vLLM-omni`

			`vLLM-omni extends vLLM capabilities to support multi-modality model inference and serving, particularly focusing on non-autoregressive structures and non-textual outputs. It aims to support:`

			`- Multi-modal Models: Text, image, video, audio, and sensor data processing`
			`- Non-autoregressive Architectures: Diffusion Transformers (DiT) and other parallel generation models`
			`- Heterogeneous Outputs: Beyond traditional text generation to multimodal outputs`

			`## <span class="twemoji">🎯</span> Key Features`

			`### Multi-Engine Support`

			`- Autoregressive Engine: Traditional text generation with enhanced KV-caching`
			`- Diffusion Engine: Support for DiT models and iterative generation`
			`- Hybrid Engine: Combined AR+DiT processing pipelines`

			`### Modality Processing`

			`- Text: Advanced tokenization and embedding generation`
			`- Image: Vision encoder integration (CLIP, etc.)`
			`- Audio: Speech processing and audio embedding`

			`## <span class="twemoji">🤖</span> Supported Models`

			`- Qwen-omni (Thinker-Talker-Codec structure)`
			`- HunyunaImage 3.0 (Ongoing)`
			`- Qwen-Image (Ongoing)`

			`## <span class="twemoji">📚</span> Documentation Navigation`

[Feature] Support online inference (#64 ) 2025-11-17 14:44:10 +08:00			`- To run open-source models on vLLM-Omni, we recommend starting with the [:material-code-tags: User Quide](user_guide/getting_started/quickstart.md)`
[Docs] Setup Documentation System and Re-organize Dependencies (#49 ) 2025-11-15 21:45:43 +08:00			`- To develop and contribute to vLLM-Omni, we recommend starting with the [:material-tools: Developer Guide](contributing/README.md)`