Summary

This repository provides custom ComfyUI nodes designed to integrate with the "Zhenzhen" API service, offering affordable access to a wide range of cutting-edge AI models for image and video generation. It targets hobbyists and fans looking to experiment with the latest AI technologies without significant upfront costs, simplifying complex model interactions through a unified API.

How It Works

The project offers a collection of ComfyUI nodes that act as wrappers for various third-party AI model APIs, primarily those offered by the associated "Zhenzhen" service. Users integrate these nodes into their ComfyUI workflows to access models like Sora2, NanoBanana, Gemini, Veo, and others for tasks such as text-to-video, image generation, and video editing, abstracting away direct API complexities.

Quick Start & Requirements

• Primary Install/Run: Integration into an existing ComfyUI environment by adding custom nodes.
• Prerequisites: A functional ComfyUI setup is required. Users must register for an API key at https://ai.t8star.cn to access the underlying services.
• Links:
  • API Service: https://ai.t8star.cn
  • Online Workflows: https://www.runninghub.ai/ (Overseas), https://www.runninghub.cn/ (Domestic)
  • Tutorials: Extensive video guides are available on Bilibili and YouTube (links provided in README).

Highlighted Details

• Extensive Model Support: Integrates a vast and frequently updated array of AI models including Sora2 (various versions), NanoBanana (Pro, 2, 4K), Gemini 3 Pro, Veo3.1, Grok Video, Flux, Qwen, MinMax, Kling, Jimeng, and Suno.
• Affordable API Access: Positions itself as a cost-effective solution for accessing advanced AI models, with detailed pricing structures for various services.
• Rapid Development: Undergoes frequent updates, with a detailed changelog indicating regular addition of new nodes, model support, bug fixes, and price adjustments.
• ComfyUI Workflow Integration: Provides specialized nodes that seamlessly fit into ComfyUI's node-based workflow paradigm, simplifying complex AI task orchestration.

Maintenance & Community

Maintenance appears active, evidenced by the extensive and recent changelog detailing frequent updates. Community engagement is primarily driven through video tutorials on Bilibili and YouTube, with specific links provided for various model demonstrations. No direct community channels like Discord or Slack are mentioned.

Licensing & Compatibility

No explicit open-source license is stated in the provided README content. The associated service explicitly states it is "not for commercial secondary development" and is "only for fans," implying significant restrictions on commercial use. Users are also responsible for their own taxes (5-25%).

Limitations & Caveats

Functionality is entirely dependent on the availability and pricing of the associated "Zhenzhen" API service. Commercial use is explicitly prohibited. Users must navigate the service's pricing structure and tax responsibilities. The project's rapid development means users may encounter evolving APIs, potential instability during updates, and reliance on external service uptime.