InvokeAI: Web UI for Stable Diffusion image generation
Node-based workflow interface for local Stable Diffusion deployment.
Learn more about InvokeAI
InvokeAI is a web application that provides a React-based UI for working with Stable Diffusion and related generative models. The system runs a locally hosted web server and implements a node-based architecture for constructing generation pipelines. It supports multiple model formats (ckpt and diffusers), various Stable Diffusion versions (1.5, 2.0, SDXL, FLUX), and includes features like in-painting, out-painting, upscaling, and embedding management. The tool is designed for local deployment on compatible hardware across Windows, macOS, and Linux systems.
Unified Canvas Implementation
Provides an integrated canvas with support for generation, in-painting, out-painting, and brush tools in a single interface. Images and metadata can be dragged and dropped across UI elements for workflow continuity.
Node-Based Workflow System
Offers a node-based workflow editor that allows users to construct and share custom generation pipelines. Workflows can be saved and reused for specific production use cases.
Multi-Model Support
Supports both checkpoint and diffusers model formats across multiple Stable Diffusion versions and FLUX models. Includes built-in model and embedding managers for organizing and switching between different models.
from invokeai.app.invocations.baseinvocation import BaseInvocation, InvocationContext
from invokeai.app.invocations.primitives import ImageOutput
class TextToImageInvocation(BaseInvocation):
prompt: str
width: int = 512
height: int = 512
def invoke(self, context: InvocationContext) -> ImageOutput:
image = context.services.generate_image(
prompt=self.prompt,
width=self.width,
height=self.height
)
return ImageOutput(image=image)Related Repositories
Discover similar tools and frameworks used by developers
Goose
LLM-powered agent automating local software engineering workflows.
MMDetection
Modular PyTorch framework for object detection research and deployment.
OpenPose
Multi-person 135-keypoint anatomical detection in C++.
YOLOv5
Real-time object detection with cross-platform deployment support.
Llama
PyTorch inference for Meta's Llama language models.