HyperChatFeatured

What is HyperChat?

HyperChat is an open-source chat client that supports the Model Context Protocol (MCP) and can use various LLM APIs to deliver an advanced chat experience and productivity tools. It runs on Windows, macOS, and Linux, and is also accessible via command line, Docker, or web browser.

How to use HyperChat?

Configure your API key (any OpenAI‑compatible LLM service), then ensure you have uv (or uvx) and nodejs (with npx) installed. Run npx -y @dadigua/hyper-chat to start a local web server on port 16100 with default password 123456. Alternatively, pull the Docker image with docker pull dadigua/hyperchat-mini:latest.

Key features of HyperChat

• Supports OpenAI, Claude, Qwen, Deepseek, Gemini, and more LLMs
• Fully supports MCP for tools, resources, and prompts
• WebDAV incremental synchronization with hash‑based speed
• HyperPrompt syntax with variables and real‑time preview
• Agents with preset prompts and permitted MCP selection
• ChatSpace for simultaneous multi‑dialogue chats

Use cases of HyperChat

• Automate repetitive tasks by defining agents with specific prompts and MCP tools
• Compare model outputs side‑by‑side in the same chat session
• Render interactive content like Artifacts, SVG, HTML, and Mermaid diagrams
• Schedule agents to run periodic tasks and monitor completion status
• Create a knowledge base using MCP‑based RAG (retrieval‑augmented generation)

FAQ from HyperChat

What platforms does HyperChat run on?

HyperChat runs on Windows, macOS, and Linux. It also offers a web interface accessible from any device with a browser.

Which LLMs are supported?

HyperChat supports all OpenAI‑style APIs, including OpenAI, Claude, Qwen, Deepseek, GLM, Ollama, xAI, Gemini, and more.

Does HyperChat support MCP?

Yes, HyperChat fully supports MCP, including Resources, Prompts, and Tools. You can install third‑party MCP servers and select individual tools.

Is HyperChat free and open source?

Yes, HyperChat is open source (see its GitHub repository). The README does not mention any paid tiers or licenses beyond its disclaimer that it is for learning and communication.

Are there any known limitations?

The README does not list specific limitations. It notes that the project is for learning and communication only and that users are responsible for their own actions when using it.