Skip to main content

OpenAI Official SDK

note

This is the documentation for the OpenAI Official SDK integration, that uses the official OpenAI Java SDK.

LangChain4j provides 4 different integrations with OpenAI for using embedding models, and this is #2 :

  • OpenAI uses a custom Java implementation of the OpenAI REST API, that works best with Quarkus (as it uses the Quarkus REST client) and Spring (as it uses Spring's RestClient).
  • OpenAI Official SDK uses the official OpenAI Java SDK.
  • Azure OpenAI uses the Azure SDK from Microsoft, and works best if you are using the Microsoft Java stack, including advanced Azure authentication mechanisms.
  • GitHub Models uses the Azure AI Inference API to access GitHub Models.

Use cases for this integration​

This integration uses the OpenAI Java SDK GitHub Repository, and will work for all OpenAI models which can be provided by:

  • OpenAI
  • Azure OpenAI
  • GitHub Models

It will also work with models supporting the OpenAI API.

OpenAI Documentation​

Maven Dependency​

<dependency>
    <groupId>dev.langchain4j</groupId>
    <artifactId>langchain4j-open-ai-official</artifactId>
    <version>1.0.0-beta2</version>
</dependency>

Configuring the models​

To use OpenAI models, you usually need an endpoint URL, an API key, and a model name. This depends on where the model is hosted, and this integration tries to make it easier with some auto-configuration:

Generic configuration​

import com.openai.models.embeddings.EmbeddingModel;
import dev.langchain4j.model.embedding.EmbeddingModel;
import dev.langchain4j.model.openaiofficial.OpenAiOfficialEmbeddingModel;

// ....

EmbeddingModel model = OpenAiOfficialEmbeddingModel.builder()
        .baseUrl(System.getenv("AZURE_OPENAI_ENDPOINT"))
        .apiKey(System.getenv("AZURE_OPENAI_KEY"))
        .modelName(EmbeddingModel.TEXT_EMBEDDING_3_SMALL)
        .build();

Specific configurations for Azure OpenAI and GitHub Models.​

Similar to configuring the OpenAI Official Chat Model, you can configure the OpenAiOfficialEmbeddingModel with Azure OpenAI and GitHub Models, using the isAzure() and isGitHubModels() methods.

For Azure OpenAI:

EmbeddingModel model = OpenAiOfficialEmbeddingModel.builder()
        .baseUrl(System.getenv("AZURE_OPENAI_ENDPOINT"))
        .apiKey(System.getenv("AZURE_OPENAI_KEY"))
        .modelName(EmbeddingModel.TEXT_EMBEDDING_3_SMALL)
        .isAzure(true) // Not necessary if the base URL ends with `openai.azure.com`
        .build();

For GitHub Models:

EmbeddingModel model = OpenAiOfficialEmbeddingModel.builder()
        .modelName(EmbeddingModel.TEXT_EMBEDDING_3_SMALL)
        .isGitHubModels(true)
        .build();

Using the models​

Once the model is configured, you can use it to create embeddings:

Response<Embedding> response = model.embed("Please embed this sentence.");