SGLang Diffusion#

SGLang Diffusion is an inference framework for accelerated image and video generation using diffusion models. It provides an end-to-end unified pipeline with optimized kernels and an efficient scheduler loop.

Key Features#

Broad Model Support: Wan series, FastWan series, Hunyuan, Qwen-Image, Qwen-Image-Edit, Flux, Z-Image, GLM-Image, and more
Fast Inference: Optimized kernels, efficient scheduler loop, and Cache-DiT acceleration
Ease of Use: OpenAI-compatible API, CLI, and Python SDK
Multi-Platform: NVIDIA GPUs (H100, H200, A100, B200, 4090) and AMD GPUs (MI300X, MI325X)

Quick Start#

Installation#

uv pip install "sglang[diffusion]" --prerelease=allow

See Installation Guide for more installation methods and ROCm-specific instructions.

Basic Usage#

Generate an image with the CLI:

sglang generate --model-path Qwen/Qwen-Image \
    --prompt "A beautiful sunset over the mountains" \
    --save-output

Or start a server with the OpenAI-compatible API:

sglang serve --model-path Qwen/Qwen-Image --port 30010

Documentation#

Getting Started#

Installation - Install SGLang Diffusion via pip, uv, Docker, or from source
Compatibility Matrix - Supported models and optimization compatibility

Usage#

CLI Documentation - Command-line interface for sglang generate and sglang serve
OpenAI API - OpenAI-compatible API for image/video generation and LoRA management

Performance Optimization#

Performance Overview - Overview of all performance optimization strategies
Attention Backends - Available attention backends (FlashAttention, SageAttention, etc.)
Caching Strategies - Cache-DiT and TeaCache acceleration
Profiling - Profiling techniques with PyTorch Profiler and Nsight Systems

Reference#

Environment Variables - Configuration via environment variables
Support New Models - Guide for adding new diffusion models
Contributing - Contribution guidelines and commit message conventions
CI Performance - Performance baseline generation script

CLI Quick Reference#

Generate (one-off generation)#

sglang generate --model-path <MODEL> --prompt "<PROMPT>" --save-output

Serve (HTTP server)#

sglang serve --model-path <MODEL> --port 30010

Enable Cache-DiT acceleration#

SGLANG_CACHE_DIT_ENABLED=true sglang generate --model-path <MODEL> --prompt "<PROMPT>"

SGLang Diffusion

Contents