Spring AI Tracing

How to use OpenInference instrumentation with Spring AI and export traces to Arize Phoenix.

Prerequisites

• Java 11 or higher
• (Optional) Phoenix API key if using auth

Add Dependencies

1. Gradle

Add the dependencies to your build.gradle:

dependencies {
	implementation 'org.springframework.ai:spring-ai-starter-model-openai'
	implementation 'io.micrometer:micrometer-tracing-bridge-brave:1.5.1'
	implementation project(path: ':instrumentation:openinference-instrumentation-springAI')

	// OpenTelemetry
	implementation "io.opentelemetry:opentelemetry-sdk"
	implementation "io.opentelemetry:opentelemetry-exporter-otlp"
	implementation "io.opentelemetry:opentelemetry-exporter-logging"

	testImplementation 'org.springframework.boot:spring-boot-starter-test'
	testRuntimeOnly 'org.junit.platform:junit-platform-launcher'
}

Setup Phoenix

Local

uvx arize-phoenix serve

No uv? pip install arize-phoenix && phoenix serve does the same thing. See Terminal setup for customization.

Container

docker run -p 6006:6006 -p 4317:4317 arizephoenix/phoenix:latest

Images are published to Docker Hub. See Docker for volumes, PostgreSQL, and other options.

Self-Host

Run Phoenix on your own infrastructure, backed by PostgreSQL so traces persist beyond a single process. This is the option to reach for once Phoenix is shared across a team or environment.

The self-hosting guide covers Kubernetes, Helm, Railway, AWS CloudFormation, Google Cloud Run, Azure, and Render, plus authentication and configuration.

Phoenix serves its UI and OTLP HTTP on port 6006, and OTLP gRPC on port 4317. For a local instance that’s http://localhost:6006. Leave it running while you work.

If your Phoenix instance isn’t running on localhost, adjust the endpoint in the code below as needed. If it has authentication enabled, also set PHOENIX_API_KEY to an API key from its Settings page.

Configuration for Phoenix Tracing

private static void initializeOpenTelemetry() {
        // Create resource with service name
        Resource resource = Resource.getDefault()
                .merge(Resource.create(Attributes.of(
                        AttributeKey.stringKey("service.name"), "spring-ai",
                        AttributeKey.stringKey(SEMRESATTRS_PROJECT_NAME), "spring-ai-project",
                        AttributeKey.stringKey("service.version"), "0.1.0")));

        String apiKey = System.getenv("PHOENIX_API_KEY");
        OtlpGrpcSpanExporterBuilder otlpExporterBuilder = OtlpGrpcSpanExporter.builder()
                .setEndpoint("http://localhost:4317") // adjust as needed
                .setTimeout(Duration.ofSeconds(2));
        OtlpGrpcSpanExporter otlpExporter = null;
        if (apiKey != null && !apiKey.isEmpty()) {
            otlpExporter = otlpExporterBuilder
                    .setHeaders(() -> Map.of("Authorization", String.format("Bearer %s", apiKey)))
                    .build();
        } else {
            logger.log(Level.WARNING, "Please set PHOENIX_API_KEY environment variable if auth is enabled.");
            otlpExporter = otlpExporterBuilder.build();
        }

        // Create tracer provider with both OTLP (for Phoenix) and console exporters
        tracerProvider = SdkTracerProvider.builder()
                .addSpanProcessor(BatchSpanProcessor.builder(otlpExporter)
                        .setScheduleDelay(Duration.ofSeconds(1))
                        .build())
                .addSpanProcessor(SimpleSpanProcessor.create(LoggingSpanExporter.create()))
                .setResource(resource)
                .build();

        // Build OpenTelemetry SDK
        OpenTelemetrySdk.builder()
                .setTracerProvider(tracerProvider)
                .setPropagators(ContextPropagators.create(W3CTraceContextPropagator.getInstance()))
                .buildAndRegisterGlobal();

        System.out.println("OpenTelemetry initialized. Traces will be sent to Phoenix at http://localhost:6006");
    }
}

Run Spring AI

By instrumenting your application, spans will be created whenever it is run and will be sent to the Phoenix server for collection.

import com.arize.instrumentation.springAI.SpringAIInstrumentor;
import org.springframework.ai.openai.OpenAiChatModel;

initializeOpenTelemetry();

// 2. Create OITracer + instrumentor
OITracer tracer = new OITracer(tracerProvider.get("com.example.springai"), TraceConfig.getDefault());
ObservationRegistry registry = ObservationRegistry.create();
registry.observationConfig().observationHandler(new SpringAIInstrumentor(tracer));

// 3. Build Spring AI model
String apiKey = System.getenv("OPENAI_API_KEY");
OpenAiApi openAiApi = OpenAiApi.builder().apiKey(apiKey).build();
OpenAiChatOptions options = OpenAiChatOptions.builder().model("gpt-5.6-luna").build();

OpenAiChatModel model = OpenAiChatModel.builder()
    .openAiApi(openAiApi)
    .defaultOptions(options)
    .observationRegistry(registry)
    .build();

// 4. Use it — traces are automatically created
ChatResponse response = model.call(new Prompt("What is the capital of France?"));
System.out.println("Response: " + response.getResult().getOutput().getContent());

Full example: https://github.com/Arize-ai/openinference/blob/main/java/examples/spring-ai-example/src/main/java/com/arize/openinference/examples/SpringAI.java

Observe

Once configured, your OpenInference traces will be automatically sent to Phoenix where you can:

• Monitor Performance: Track latency, throughput, and error rates
• Analyze Usage: View token usage, model performance, and cost metrics
• Debug Issues: Trace request flows and identify bottlenecks
• Evaluate Quality: Run evaluations on your LLM outputs

Resources

Full Example

OpenInference package