Java 11 and newer ship java.net.http.HttpClient, so you don't need extra dependencies for a basic client. This example uses Jackson for JSON, which most projects already have.
import com.fasterxml.jackson.databind.*;
import java.net.URI;
import java.net.http.*;
import java.time.Duration;
import java.util.*;
public class Ai {
private static final HttpClient http = HttpClient.newBuilder().connectTimeout(Duration.ofSeconds(20)).build();
private static final ObjectMapper json = new ObjectMapper();
public static String chat(String model, String prompt) throws Exception {
Map<String, Object> body = Map.of(
"model", model,
"max_tokens", 800,
"messages", List.of(Map.of("role", "user", "content", prompt))
);
HttpRequest req = HttpRequest.newBuilder(URI.create("https://devsaio.lol/v1/chat/completions"))
.timeout(Duration.ofSeconds(120))
.header("Authorization", "Bearer " + System.getenv("DEVSAIO_API_KEY"))
.header("Content-Type", "application/json")
.POST(HttpRequest.BodyPublishers.ofString(json.writeValueAsString(body)))
.build();
HttpResponse<String> res = http.send(req, HttpResponse.BodyHandlers.ofString());
JsonNode root = json.readTree(res.body());
if (res.statusCode() != 200) throw new RuntimeException(root.path("error").path("message").asText("AI error " + res.statusCode()));
return root.path("choices").get(0).path("message").path("content").asText();
}
public static void main(String[] args) throws Exception {
System.out.println(chat("claude-sonnet-4-6", "Give me 3 tips for writing clean Java."));
}
}Spring Boot
Spring AI's OpenAI integration accepts a custom base URL in configuration (spring.ai.openai.base-url). Point it at https://devsaio.lol or https://devsaio.lol/v1 depending on whether your Spring AI version appends /v1 itself, set the API key and choose a model in spring.ai.openai.chat.options.model.
Claude, GPT, Gemini, DeepSeek, image and video models through one OpenAI- and Anthropic-compatible endpoint. Prepaid, pay only for what you use.