AI assistant kit with Chinese payment rails and multi-model support
A turnkey open-source kit for running a multi-model AI chat service with built-in image generation, music synthesis, and native Chinese payment rails.

What it does
GeekAI is a full-stack, deployable AI assistant platform built in Go and Vue 3. It bundles a chat frontend, an admin dashboard, and backend integrations for multiple large language models—OpenAI, Claude, Tongyi Qianwen, Kimi, DeepSeek, and Gitee AI—alongside MidJourney and Stable Diffusion for image generation and Suno for text-to-music. The project essentially packages the infrastructure needed to run a small AI SaaS, including membership tiers, point-card billing, and role-based chat personas, into a single codebase.
The interesting bit
The project is unapologetically aimed at commercial operators in the Chinese market: it supports personal WeChat QR codes as payment channels to bypass enterprise merchant requirements, and it opens with a legal warning about unregistered generative-AI services. That focus on turnkey monetization—plus pre-baked personas like Confucius and Steve Jobs—makes it feel less like a generic chat UI and more like a vending machine you can skin and deploy.
Key highlights
- Bundles LLM chat, MidJourney/Stable Diffusion image generation (text-to-image, image-to-image, face swap, fusion), and Suno music synthesis behind one WebSocket-based interface.
- Ships with pre-configured personas (Socrates, weekly-report bot, Xiaohongshu writer) and a plugin API that uses LLM function calling to pull live data such as Weibo hot search and Toutiao news.
- Targets solo operators with built-in membership tiers, point-card billing, and the ability to accept payments via personal WeChat QR codes or Alipay.
- Provides native desktop and mobile clients for Windows, Linux, macOS, and Android.
- A superset fork, GeekAI-PLUS, adds DALL-E, text-to-mind-map generation, and per-model API key routing.
Caveats
- The feature list claims WeChat Pay is integrated, but the project todo list still marks it as unfinished, so the exact payment state is unclear.
- Knowledge-base Q&A (RAG) and text-to-video generation remain on the roadmap and are not yet implemented.
- The author explicitly refuses bug reports via WeChat or social channels, requiring all issues to be filed on GitHub.
Verdict
Worth exploring if you want to launch a multi-model AI service with Chinese payment integration and out-of-the-box image generation. Look elsewhere if you need deep retrieval-augmented generation or a billing stack tailored for Western payment processors.
Frequently asked
- What is yangjian102621/geekai?
- A turnkey open-source kit for running a multi-model AI chat service with built-in image generation, music synthesis, and native Chinese payment rails.
- Is geekai open source?
- Yes — yangjian102621/geekai is open source, released under the Apache-2.0 license.
- What language is geekai written in?
- yangjian102621/geekai is primarily written in Vue.
- How popular is geekai?
- yangjian102621/geekai has 4.7k stars on GitHub.
- Where can I find geekai?
- yangjian102621/geekai is on GitHub at https://github.com/yangjian102621/geekai.