This repository provides ComfyUI nodes to integrate Stable Diffusion 3 (SD3) and SD3 Turbo models via the Stability AI API. It enables text-to-image and image-to-image generation with various aspect ratios, targeting ComfyUI users who want to leverage the latest SD3 capabilities without local model hosting.

How It Works

The nodes act as a client for the Stability AI API, abstracting the API calls for SD3 and SD3 Turbo. Users provide a Stability API key and select generation parameters like prompt, negative prompt (for SD3), aspect ratio, model type, and seed. For image-to-image, an input image and strength parameter are also supported. This approach offloads the computational burden to Stability AI's infrastructure, allowing users to access powerful models with minimal local setup.

Quick Start & Requirements

• Install via git clone https://github.com/ZHO-ZHO-ZHO/ComfyUI-StableDiffusion3-API into custom_nodes.
• Requires a Stability API key, with 25 free credits per account. SD3 generation costs 6.5 credits/image, SD3 Turbo costs 4 credits/image.
• Add API key to config.json.
• ComfyUI Manager installation is planned.
• Official documentation and demo links are not provided in the README.

Highlighted Details

• Supports SD3 and SD3 Turbo models.
• Offers 9 aspect ratio options for text-to-image.
• Includes image-to-image functionality with strength control.
• Fixes for image-to-image and Turbo bugs in V1.5.

Maintenance & Community

• Contact: zhozho3965@gmail.com
• QQ Group: 839821928
• Social media links provided for Bilibili and X (Twitter).
• Support options via Bilibili and "爱发电".

Licensing & Compatibility

• The repository does not explicitly state a license.
• Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The project is dependent on the Stability AI API, which may incur costs and has usage limits. The README mentions that image-to-image and Turbo features had issues in V1.0, though V1.5 claims to have fixed these.