HiveChat

What is HiveChat?

HiveChat is an AI chat application designed for small and medium teams. It supports multiple large language models such as Deepseek, OpenAI, Claude, and Gemini, and allows configuring MCP servers via SSE mode. The application provides both an end-user chat interface and an admin panel for team management.

How to use HiveChat?

End users simply log in and start chatting. For administrators, the application requires installing dependencies (npm install), configuring a .env file with a PostgreSQL database URL and authentication secrets, initializing the database (npm run initdb), and then building and starting the server. Alternatively, it can be deployed via Docker or on Vercel with a one-click button. After deployment, the admin sets up the first admin account by visiting /setup.

Key features of HiveChat

• Admins configure models once for the whole team
• Supports email, WeCom, DingTalk, and Feishu login
• Group-based user management with per-group model access and monthly token limits
• DeepSeek chain-of-thought display, LaTeX/Markdown rendering, image understanding
• AI agents and cloud data storage
• Supports 14+ model providers plus custom OpenAI-style providers

Use cases of HiveChat

• Small companies or schools providing a unified AI chat interface for team members
• Teams needing centralized control of model access and token consumption per group
• Organizations that require third-party login integration (WeCom, DingTalk, Feishu) for authentication
• Deploying an on-premises AI chat solution with MCP server connectivity

FAQ from HiveChat

What databases does HiveChat require?

HiveChat requires a PostgreSQL database. For local deployment, you must install PostgreSQL or connect to a remote one; Vercel deployment can use Neon (Serverless Postgres) provided by the Vercel platform.

How do I update the database schema after upgrading?

For npm-based installations, run npm run initdb to apply any schema changes—it is safe to run even if no updates exist. For Docker upgrades, manual SQL may be required (see the release notes for specific versions).

Does HiveChat support public registration?

Yes, the admin can enable or disable account registration in the admin panel, making it suitable for closed teams like companies or schools.

How is data stored?

Data is stored in a PostgreSQL database. The application supports cloud data storage (via the configured database) but does not rely on any external cloud service except the database.

What transport does MCP server configuration use?

MCP servers can be configured using SSE (Server-Sent Events) mode only, as stated in the feature list.