This project provides a minimal viable product (MVP) web frontend for AI chatbots, targeting developers who need a quick start for building large language model (LLM) applications. It offers a clean, modular structure and integrates with various LLM APIs, simplifying the development of chat interfaces with features like streaming output and markdown rendering.

How It Works

The project leverages a modern Vue 3, Vite 6, and TypeScript stack, with Naive UI for components and Pinia for state management. It employs a single-turn conversation model, meaning each user query is processed independently without maintaining chat history. The architecture is designed for flexibility, allowing easy integration of different LLM APIs through a unified mapping mechanism and response transformation functions.

Quick Start & Requirements

• Install dependencies: pnpm i
• Run locally: pnpm dev
• Prerequisites: Node.js >= 18.12.x, pnpm 9.x.
• Configuration: Copy .env.template to .env and fill in API keys for desired LLM services.
• Live Demo: [Link to live demo not provided in README]

Highlighted Details

• Supports multiple LLM APIs including DeepSeek, Spark, Kimi Moonshot, SiliconFlow, and Ollama.
• Features markdown rendering with code highlighting, Mermaid diagrams, and KaTex/LaTeX math support.
• Includes a local mock data mode for development without API keys.
• Utilizes Vite's proxy for handling cross-origin requests during development.

Maintenance & Community

• The project is actively maintained by pdsuwwz.
• Community contributions are welcomed via issues and pull requests.
• Related projects are listed in the README.

Licensing & Compatibility

• Licensed under the MIT License.
• Permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

The current implementation strictly supports single-turn conversations and does not retain chat history, though future multi-turn support is mentioned as a possibility without a concrete plan. The proxy configuration is only for development environments; production deployment requires separate CORS or reverse proxy setup.