qwen-image

SGLang is a high-performance serving framework for large language models and multimodal models.

🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.

Diffusion model(SD,Flux,Wan,Qwen Image,Z-Image,...) inference in pure C/C++

Lightweight Image Video Action Generation Inference Framework

Apple MLX native implementations of state-of-the-art generative image & video models

Open-source SOTA multi-image editing model

A Collection of Google Colab Notebooks for various projects

Create high-fidelity Codex companions from 2+ photos with local open-weight neural editing and no OpenAI API key.

Qwen Image models through MPS

Qwen Image Integrated KSampler, intelligent multimodal sampler, support Z-Image, optimizes official offset issues, image scaling, multiple reference images, automatic VRAM/RAM management, batch generation, automatic saving, sound notification / 千问图像集成采样器 - K采样器，支持z-image，优化官方偏移问题，图片缩放、多张参考图、自动显存/内存管理、批量生成、自动保存、声音通知、AuraFlow优化、CFG标准化调节等功能

ComfyUI Nodes for AsymFlow and pi-Flow

all ComfyUI http/ws APIs. support NodeJS / Browser environments. and advanced programmable workflow.

open source style transfer model on par with nano banana pro, supporting Qwen-Image-Edit 2509, 2511, SenseNovaU1

The most complete, up-to-date comparison of AI image generation models — which model, via which API, at what price.

WeeLLM runs large diffusion models with as little as 4 GB of VRAM, without any quantization. It dynamically determines how many layers can fit within the available VRAM and streams the text encoder and transformer layers to the GPU layer by layer, enabling inference on hardware with limited VRAM. It supports both safetensors and GGUF models.

KsanaDiT: High-Performance DiT (Diffusion Transformer) Inference Framework for Video & Image Generation

集成了众多功能的API节点，只需要一个API-key即可以调用全网的API接口能力，低配电脑也能顺利运行

[ICML 2026] Elastic Diffusion Transformer: Accelerating SOTA generation models (e.g., Qwen-Image, Hunyuan3d ) through adaptive computation

[CVPR 2026] VDE: Training-Free Accelerating Rectified Flow Model via Velocity Decomposition and Estimation

The first model supporting content-preserving style transfer for Qwen-Image