Spring Ai Mcp Deepseek

What is Spring Ai Mcp Deepseek?

Spring Ai Mcp Deepseek is a project that uses Spring AI 1.0.0 to integrate Model Context Protocol (MCP) services with a DeepSeek AI client. It provides an MCP server and multiple client modules for chat and tool invocation.

How to use Spring Ai Mcp Deepseek?

Configure your DeepSeek API key (spring.ai.openai.api-key) in application.properties. Start the mcp-server module on port 8085, then start either the mcp-client module (port 8082) for a text chat API, or the mcp-tool-client module (port 8083) for a page-based client with stream and tool‑call control. Access the SSE endpoint at http://localhost:8085/sse.

Key features of Spring Ai Mcp Deepseek

• MCP server with SSE streaming endpoint on port 8085
• Text chat client API on port 8082
• Page client with call/stream and tool control on port 8083
• Displays registered MCP tools on the client page
• Formats tool call results for readability

Use cases of Spring Ai Mcp Deepseek

• Chat with DeepSeek AI via an MCP‑integrated text API
• Invoke MCP tools from a web page client
• Stream AI responses in real time
• Test and demonstrate MCP tool‑calling capabilities

FAQ from Spring Ai Mcp Deepseek

What API key is required?

You need a DeepSeek API key, configured as spring.ai.openai.api-key in the application.properties files of the mcp-client and call-mcp modules.

What modules does the project contain?

It contains three modules: mcp-server (port 8085), mcp-client (port 8082), and mcp-tool-client (port 8083).

How do I access the client interfaces?

For the text client, use endpoints like /api/chat or /chat with a message parameter. For the page client, visit http://localhost:8083/.

Does the server support SSE streaming?

Yes, the MCP server exposes an SSE endpoint at http://localhost:8085/sse.

Is there a user interface for tool control?

Yes, the mcp-tool-client module provides a web page at port 8083 that shows a list of registered MCP tools and allows invoking them with formatted results.