agents-flex/docs/intro/what-is-agentsflex.md

What is Agents-Flex?

Agents-Flex is an AI application development framework developed in Java, designed to simplify AI application development. It draws inspiration from LangChain, LlamaIndex, and the author's best practices as a frontline AI application developer, providing API support across AI service providers that is portable and not limited to Java development frameworks.

Agents-Flex is suitable for various scenarios, including chat, image generation, embedding models, function calling, and RAG applications, and supports both synchronous and streaming API options.

Comparison between Agents-Flex and other frameworks

1、More universally applicable

Compared to Spring-AI and LangChain4j, Agents-Flex is more universally applicable.

1. For example, Spring-AI requires JDK version JDK 21+ whereas Agents-Flex only needs JDK8+.
2. Spring-AI requires usage within the Spring framework, whereas Agents-Flex supports integration with any framework and provides spring-boot-starter.

2、Simpler API design

With Agents-Flex, chat functionality can be implemented in just two lines of code.

@Test
public void testChat() {
    OpenAiLlm llm = new OpenAiLlm.of("sk-rts5NF6n*******");
    String response = llm.chat("what is your name?");

    System.out.println(response);
}

Function Calling also requires just a few lines of code with Agents-Flex.

public class WeatherUtil {

    @FunctionDef(name = "get_the_weather_info", description = "get the weather info")
    public static String getWeatherInfo(
        @FunctionParam(name = "city", description = "the city name")String name ) {
        //Here, we should retrieve API information through third-party interfaces
        return name + "weather is cloudy with overcast. ";
    }


    public static void main(String[] args) {
        OpenAiLlm llm = new OpenAiLlm.of("sk-rts5NF6n*******");

        FunctionPrompt prompt = new FunctionPrompt("What's the weather like in Beijing today?", WeatherUtil.class);
        FunctionResultResponse response = llm.chat(prompt);

        Object result = response.getFunctionResult();

        System.out.println(result);
        //"The weather in Beijing is overcast turning to cloudy. "
    }
}

2、More Powerful Agents Orchestration

We know that a powerful AI application often requires flexible orchestration capabilities. Compared to Agents-Flex, Spring-AI and LangChain4j lack almost any orchestration capabilities.

Below is a simple example code of Agents-Flex regarding Chain (execution chain) orchestration:

public static void main(String[] args) {
    SequentialChain ioChain1 = new SequentialChain();
    ioChain1.addNode(new Agent1("agent1"));
    ioChain1.addNode(new Agent2("agent2"));

    SequentialChain ioChain2 = new SequentialChain();
    ioChain2.addNode(new Agent1("agent3"));
    ioChain2.addNode(new Agent2("agent4"));
    ioChain2.addNode(ioChain1);

    Object result = ioChain2.executeForResult("your params");
    System.out.println(result);
}

The above code implements Agents orchestration as shown in the diagram below:

The data flow is as follows: agent3 --> agent4 --> chain1, and within chain1, there is the process of agent1 --> agent2.

In Agents-Flex, we have built-in three different types of Agents execution chains:

• SequentialChain: Executes agents sequentially.
• ParallelChain: Executes agents concurrently (in parallel).
• LoopChain: Executes agents in a loop.

Moreover, each of these three chains can serve as a sub-chain for other chains, thus forming powerful and complex Agents chains.