Solon AI

v1.0.0Coding Agentsstable

Java AI application development framework (supports LLM-tool,skill; RAG; MCP; Agent-ReAct,Team-Agent). Compatible with java8 ~ java25. It can also be embedded in SpringBoot, jFinal, Vert.x, Quarkus, and other frameworks.

aichatdeepseekembeddingfunction-call
Share:
388
Stars
0
Downloads
0
Weekly
0/5

What is Solon AI?

Solon AI is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to java ai application development framework (supports llm-tool,skill; rag; mcp; agent-react,team-agent). compatible with java8 ~ java25. it can also be embedded in springboot, jfinal, vert.x, quarkus, a...

Java AI application development framework (supports LLM-tool,skill; RAG; MCP; Agent-ReAct,Team-Agent). Compatible with java8 ~ java25. It can also be embedded in SpringBoot, jFinal, Vert.x, Quarkus, and other frameworks.

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Java AI application development framework (supports LLM-tool

Use Cases

Java AI framework
LLM tool and RAG support
SpringBoot integration
opensolon

Maintainer

LicenseApache-2.0
Languagejava
Versionv1.0.0
UpdatedMay 21, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx solon-ai

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Solon AI

Solon AI is a Java-based AI application development framework that provides a unified API for LLM calls, RAG pipelines, MCP server/client integration, and multi-agent coordination (ReAct and Team-Agent patterns). It supports JDK 8 through 25 and can be embedded in SpringBoot, jFinal, Vert.x, Quarkus, and other Java frameworks via simple Maven or Gradle dependencies. Java developers use Solon AI to build production AI applications without leaving the JVM ecosystem, connecting to providers like OpenAI, Claude, DeepSeek, Ollama, and Alibaba Dashscope.

Prerequisites

  • Java 8 or later (JDK 8, 11, 17, 21, or 25 all supported)
  • Maven or Gradle build tool
  • API key for your chosen LLM provider (OpenAI, Anthropic, DeepSeek, etc.) or a running Ollama instance for local models
  • Familiarity with Java application development
1

Add the Maven dependency

Add solon-ai to your pom.xml. For Gradle projects, use the equivalent implementation declaration. This single dependency brings in the core chat, RAG, and MCP framework.

<dependency>
    <groupId>org.noear</groupId>
    <artifactId>solon-ai</artifactId>
    <version>3.x.x</version>
</dependency>
2

Configure a ChatModel

Create a ChatModel pointing at your LLM provider endpoint. The provider parameter selects the dialect adapter (openai, ollama, claude, deepseek, dashscope, etc.).

ChatModel chatModel = ChatModel.of("https://api.openai.com/v1/chat/completions")
    .provider("openai")
    .apiKey(System.getenv("OPENAI_API_KEY"))
    .model("gpt-4o")
    .build();

AssistantMessage result = chatModel.prompt("Hello").call().getMessage();
System.out.println(result.getContent());
3

Build an MCP server endpoint

Annotate a Java class with @McpServerEndpoint to expose it as an MCP server. Use @ToolMapping on methods to register individual tools that LLM clients can call.

@McpServerEndpoint(channel = McpChannel.STREAMABLE, mcpEndpoint = "/mcp")
public class WeatherMcpServer {
    @ToolMapping(description = "Get the current weather for a city")
    public String getWeather(@Param(description = "City name") String location) {
        return "It is sunny and 25°C in " + location;
    }
}
4

Connect an MCP client

Use McpClientProvider to connect your Java application to any MCP server (including the one you just built) and make its tools available to your ChatModel.

McpClientProvider mcpClient = McpClientProvider.builder()
    .channel(McpChannel.STREAMABLE)
    .url("http://localhost:8080/mcp")
    .build();

AssistantMessage result = chatModel
    .prompt("What is the weather in Tokyo?")
    .toolsAdd(mcpClient)
    .call()
    .getMessage();
5

Set up a RAG pipeline

Use Solon AI's document loading, splitting, and embedding APIs to build a retrieval-augmented generation pipeline over your own documents.

EmbeddingModel embeddingModel = EmbeddingModel.of("https://api.openai.com/v1/embeddings")
    .provider("openai")
    .apiKey(System.getenv("OPENAI_API_KEY"))
    .model("text-embedding-3-small")
    .build();

// Load and embed documents, then query with semantic retrieval

Solon AI Examples

Client configuration

Solon AI is a Java library, not a standalone MCP server process. To use a Solon AI MCP server from Claude Desktop, point the config at the running Java server's HTTP endpoint using a curl or mcp-remote bridge.

{
  "mcpServers": {
    "solon-ai": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "http://localhost:8080/mcp"]
    }
  }
}

Prompts to try

Once your Solon AI MCP server is running and connected, these prompts demonstrate what the tools can do.

- "What is the current weather in Shanghai?"
- "Search the knowledge base for information about our Q3 financial report"
- "Use the available tools to answer: what are the top 3 trending topics on our internal wiki?"
- "Call the product inventory tool and tell me how many units of SKU-1234 are in stock"
- "Summarize the documents in the /reports folder using RAG retrieval"

Troubleshooting Solon AI

ClassNotFoundException or dependency resolution failures with solon-ai

Ensure you are using a compatible JDK version (8, 11, 17, 21, or 25). Check the Solon AI GitHub releases for the latest version number and update your Maven/Gradle dependency accordingly. See solon.noear.org for the current stable version.

LLM provider calls fail with 401 Unauthorized

Confirm that your API key is set in the environment variable referenced in your code (e.g., System.getenv("OPENAI_API_KEY")). When running in a framework like SpringBoot, prefer injecting the key via application.properties rather than hardcoding it.

MCP client cannot connect to the server endpoint

Ensure your Spring/Solon application is fully started and listening on the expected port before connecting. Check that the mcpEndpoint path in @McpServerEndpoint matches the URL path used in McpClientProvider.url().

Frequently Asked Questions about Solon AI

What is Solon AI?

Solon AI is a Model Context Protocol (MCP) server that java ai application development framework (supports llm-tool,skill; rag; mcp; agent-react,team-agent). compatible with java8 ~ java25. it can also be embedded in springboot, jfinal, vert.x, quarkus, and other frameworks. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Solon AI?

Follow the installation instructions on the Solon AI GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Solon AI?

Solon AI works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Solon AI free to use?

Yes, Solon AI is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "solon-ai": { "command": "npx", "args": ["-y", "solon-ai"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Solon AI?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides