Skip to content
Glitch

Spring AI

Spring AI integrates LLMs into enterprise systems. Wrap the underlying LLM client via Spring configuration to normalize and score combined inputs before generation.

Configuration

Section titled “Configuration”
import org.springframework.ai.openai.OpenAiChatClient;
import org.springframework.ai.openai.OpenAiChatOptions;
import org.springframework.ai.openai.api.OpenAiApi;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;


@Configuration
public class GlitchConfig {


    // Bean: Creates the OpenAI API client configured to use Glitch
    @Bean
    public OpenAiApi openAiApi() {
        return new OpenAiApi(
            "https://api.golabrat.ai/v1",  // Glitch base URL
            System.getenv("GLITCH_API_KEY"),  // Your Glitch API key
            null  // Uses default HTTP client
        );
    }


    // Bean: Creates the chat client that uses the Glitch-configured API
    @Bean
    public OpenAiChatClient openAiChatClient(OpenAiApi openAiApi) {
        return new OpenAiChatClient(openAiApi, OpenAiChatOptions.builder()
            .withModel("gpt-4")
            .build());
    }
}

Basic Example

Section titled “Basic Example”
import org.springframework.ai.chat.ChatClient;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Service;


@Service
public class ChatService {


    @Autowired
    private ChatClient chatClient;


    public String chat(String userMessage) {
        // All LLM calls are automatically secured via Glitch
        return chatClient.call(
            new Prompt(userMessage)
        ).getResult().getOutput().getContent();
    }
}

Pre-screening Content with /v1/detect

Section titled “Pre-screening Content with /v1/detect”

For pre-screening content before sending to Spring AI, or when you need structured detection results, use the /v1/detect endpoint:

import org.springframework.web.client.RestClient;
import org.springframework.http.ResponseEntity;
import java.util.Map;


@Configuration
public class GlitchRestClientConfig {


    @Bean
    public RestClient glitchRestClient() {
        return RestClient.builder()
            .baseUrl("https://api.golabrat.ai/v1")
            .defaultHeader("Authorization", "Bearer " + System.getenv("GLITCH_API_KEY"))
            .build();
    }


    public boolean preScreenContent(String content) {
        Map<String, Object> requestBody = Map.of(
            "content", content,
            "direction", "input"
        );


        ResponseEntity<Map> response = glitchRestClient().post()
            .uri("/detect")
            .body(requestBody)
            .retrieve()
            .toEntity(Map.class);


        Map<String, Object> result = response.getBody();
        Boolean blocked = (Boolean) result.get("blocked");


        return Boolean.TRUE.equals(blocked);
    }
}

Error Handling

Section titled “Error Handling”

Handle security blocks in your service layer:

import org.springframework.ai.chat.ChatResponse;
import org.springframework.http.HttpStatus;
import org.springframework.web.server.ResponseStatusException;


@Service
public class ChatService {


    @Autowired
    private ChatClient chatClient;


    public String chat(String userMessage) {
        try {
            ChatResponse response = chatClient.call(new Prompt(userMessage));
            return response.getResult().getOutput().getContent();
        } catch (Exception e) {
            // Check if it's a security block (403)
            if (e.getMessage() != null && e.getMessage().contains("403")) {
                throw new ResponseStatusException(
                    HttpStatus.FORBIDDEN,
                    "Request blocked by security policy"
                );
            }
            throw e;
        }
    }
}

Next Steps

Section titled “Next Steps”