HiveChat: An AI Chat Bot for Teams Supporting Multiple LLMs

Introduction

HiveChat is an advanced AI chat application specifically designed for small and medium-sized teams, offering a centralized platform to interact with a wide array of large language models. It supports popular models such as Deepseek, OpenAI, Claude, and Gemini, alongside many others like Moonshot, VolcEngine, Alibaba Cloud, Baidu Qianfan, Tencent Hunyuan, Zhipu, Open Router, Grok, Ollama, and SiliconFlow. Built with a modern tech stack including Next.js, Tailwindcss, Auth.js, PostgreSQL, Drizzle ORM, and Ant Design, HiveChat provides a robust and scalable solution for team-based AI interactions.

Key features include:

• Comprehensive Model Support: Access and manage numerous AI models from a single interface.
• Flexible Authentication: Supports email login, WeChat Work, DingTalk, and Feishu for seamless team integration.
• User and Group Management: Administrators can create user groups, assign specific models, and set monthly token limits for each group, ensuring efficient resource allocation and cost control.
• Advanced AI Capabilities: Features include DeepSeek thought chain display, LaTeX and Markdown rendering, image understanding, and AI agents.
• Cloud Data Storage: Ensures your conversation data is securely stored.
• MCP Server Support: Integration with MCP servers using SSE mode.

Installation

HiveChat offers flexible deployment options to suit various team needs, including local setup, Docker, and Vercel.

Local Deployment

To get started locally, clone the repository, install dependencies with npm install, and configure your .env file with database details, AUTH_SECRET, and ADMIN_CODE. After setting up, run npm run initdb to initialize the database and npm run dev (for development) or npm run build followed by npm run start (for production). Finally, access /setup to initialize the administrator account.

Docker Deployment

For containerized deployment, clone the repository, copy .env.example to .env, and configure AUTH_SECRET and ADMIN_CODE. Then, simply run docker compose up -d to start the services. The administrator account can be initialized by visiting /setup in your browser.

Vercel Deployment

HiveChat can also be deployed directly to Vercel. Click the "Deploy with Vercel" button on the GitHub repository, clone the code to your GitHub, and fill in the necessary environment variables, including DATABASE_URL (which can be linked to a Neon Serverless Postgres instance), AUTH_SECRET, and ADMIN_CODE. After deployment, initialize the administrator account via the /setup path on your Vercel domain.

For detailed instructions and troubleshooting, please refer to the official GitHub README.

Examples

Experience HiveChat firsthand through its online demonstration instances:

• Cloud Version: A multi-tenant environment where you can register and explore: https://www.hivechat.net/
• Private Deployment User Client: Register and interact with the AI models: https://chat.yotuku.cn/
• Private Deployment Admin Panel: Explore the administrative features, including user and model management, using the provided demo credentials (Email: admin@demo.com, Password: helloHivechat): https://hivechat-demo.vercel.app/

These demos provide a comprehensive look at both the user-facing chat interface and the powerful backend administration tools.

Why Use HiveChat?

HiveChat stands out as an ideal solution for teams seeking to integrate AI into their workflows efficiently and securely. Its primary advantages include:

• Centralized AI Access: Provides a single, managed interface for diverse AI models, simplifying access for all team members.
• Cost Management: Granular control over token limits per user group helps manage and optimize AI usage costs.
• Scalability and Flexibility: Supports a wide range of LLMs and offers multiple deployment options, from local to cloud, adapting to your infrastructure needs.
• Enhanced Collaboration: Features like user grouping and shared access to AI tools foster better team collaboration and knowledge sharing.
• Robust Administration: A dedicated admin panel allows for easy user, model, and authentication management, ensuring secure and controlled access.
• Modern and Maintainable: Built with a contemporary tech stack, HiveChat is designed for performance and ease of maintenance.

Links

• GitHub Repository: https://github.com/HiveNexus/HiveChat
• Cloud Demo: https://www.hivechat.net/
• Private Deployment User Demo: https://chat.yotuku.cn/
• Private Deployment Admin Demo: https://hivechat-demo.vercel.app/