Springboot 🍃 + Langchain4j 🐦 + OpenAI 🤖 + GraalVM JDK 21 🚀

Build your own chatGPT inside a native image? Yes you can!

This sample app is part of a Java GenAI use case with OpenAI GPT LLM and it's part of a bigger personal project which look have multiple java samples developed with most popular Java AI frameworks (Spring AI, LangChain4j, Semantic Kernel, etc to developing amazing GenAI app using Java language.

The frontend was built on angular. Here is the repo.

Run it

Add your OpenAI API Key

It's recommended create an environment var like this:

$ export OPENAI_API_KEY=********

Otherwise you can add it in the application properties.

openai.api.key=MyOpenAIApiKey

Using the JVM

Make sure you're using GraalVM CE JDK 21

$ java -version
openjdk version "21" 2023-09-19
OpenJDK Runtime Environment GraalVM CE 21+35.1 (build 21+35-jvmci-23.1-b15)
OpenJDK 64-Bit Server VM GraalVM CE 21+35.1 (build 21+35-jvmci-23.1-b15, mixed mode, sharing)

Run springboot app

$ mvn clean spring-boot:run

Check your running app mode calling to getNativeImage API endpoint

## Send a first prompt
$ curl http://localhost:8080/openai/build-type

Using AOT compilation and generate native imnage

First, create the native image config files using the agentlib. The project already have configured it in the pom.xml, Just run it like this:

$ mvn clean spring-boot:run

Test tha API endpoint. You can use curl as well

## Send a first prompt
$ curl -X POST -H "Content-type: application/json" -d '{"message":"Hi, my name is Carlos Zela"}' http://localhost:8080/openai

## Request event-stream responses
$ curl http://localhost:8080/openai/chat-stream -v

Copy the below files from target/native-image folder to src/main/resources/META-INF/native-image folder.

• proxy-config.json
• reflect-config.json
• resource-config.json
• serialization-config.json

Build the native image

$ mvn -Pnative native:compile

Run the naitve image

$ ./target/springboot-langchain4j-graal

Test the API endpoint again!