> ## Documentation Index > Fetch the complete documentation index at: https://arize-ax.mintlify.dev/docs/llms.txt > Use this file to discover all available pages before exploring further. # Vertex AI > Trace Google Vertex AI calls (Gemini and other models) with OpenInference and send spans to Arize AX for LLM observability. [Google Vertex AI](https://cloud.google.com/vertex-ai) hosts the Gemini family of models on Google Cloud infrastructure with IAM-based auth and project-level billing. Arize AX captures every Vertex call — chat completions, tool calls, and token usage — via the [`openinference-instrumentation-google-genai`](https://github.com/Arize-ai/openinference/tree/main/python/instrumentation/openinference-instrumentation-google-genai) package. The same instrumentor that covers the Gemini API also covers Vertex — only the credential setup differs. The legacy `vertexai` Python SDK (in `google-cloud-aiplatform`) is deprecated as of June 24, 2025 with removal scheduled for June 24, 2026. This guide uses the canonical replacement: the [Google Gen AI Python SDK](https://github.com/googleapis/python-genai) (`google-genai`) configured against Vertex. ## Prerequisites * Python 3.10+ * An Arize AX account ([sign up](https://arize.com/sign-up/)) * A Google Cloud project with the [Vertex AI API enabled](https://console.cloud.google.com/apis/library/aiplatform.googleapis.com) * A service account or user with the `roles/aiplatform.user` IAM role * Authenticated Application Default Credentials (`gcloud auth application-default login`) or a service account JSON file referenced by `GOOGLE_APPLICATION_CREDENTIALS` ## Launch Arize AX 1. Sign in to your [Arize AX account](https://app.arize.com/). 2. From **Space Settings**, copy your **Space ID** and **API Key**. You will set them as `ARIZE_SPACE_ID` and `ARIZE_API_KEY` below. ## Install ```bash theme={null} pip install arize-otel openinference-instrumentation-google-genai google-genai ``` `google-genai` includes `google-auth` as a transitive dependency, so Application Default Credentials work out of the box — no extra install required. ## Configure credentials Vertex AI uses Google Cloud auth (IAM), not an API key. Authenticate locally and tell the SDK which project / region to target: ```bash theme={null} # Recommended for local dev — uses your gcloud user credentials. gcloud auth application-default login # Or, with a service account JSON file: # export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json" export ARIZE_SPACE_ID="" export ARIZE_API_KEY="" export ARIZE_PROJECT_NAME="vertexai-tracing-example" export GOOGLE_CLOUD_PROJECT="" export GOOGLE_CLOUD_LOCATION="us-central1" # optional; the SDK defaults to us-central1 ``` `GOOGLE_CLOUD_PROJECT` is mandatory. `GOOGLE_CLOUD_LOCATION` is optional and defaults to `us-central1`; set it explicitly when you need a different region (e.g. `europe-west1` for EU residency, or to match where the target model is enabled). ## Setup tracing ```python theme={null} # instrumentation.py import os from arize.otel import register from openinference.instrumentation.google_genai import GoogleGenAIInstrumentor tracer_provider = register( space_id=os.environ["ARIZE_SPACE_ID"], api_key=os.environ["ARIZE_API_KEY"], project_name=os.environ["ARIZE_PROJECT_NAME"], ) GoogleGenAIInstrumentor().instrument(tracer_provider=tracer_provider) print("Arize AX tracing initialized for Vertex AI.") ``` ## Run Vertex AI ```python theme={null} # example.py # Importing instrumentation first ensures tracing is set up # before `google.genai` is imported. from instrumentation import tracer_provider from google import genai # `vertexai=True` flips the client into Vertex mode. Project and location # come from GOOGLE_CLOUD_PROJECT / GOOGLE_CLOUD_LOCATION env vars, or pass # them explicitly: genai.Client(vertexai=True, project="...", location="...") client = genai.Client(vertexai=True) response = client.models.generate_content( model="gemini-2.5-flash", contents="Why is the ocean salty? Answer in two sentences.", ) print(response.text) ``` ### Expected output ```text wrap theme={null} Arize AX tracing initialized for Vertex AI. The ocean is salty because rivers continuously dissolve mineral salts from rocks and soil and carry them to the sea, where they accumulate over millions of years. Water leaves the ocean through evaporation but the salts remain, steadily concentrating until reaching today's roughly 3.5% salinity. ``` ## Verify in Arize AX 1. Open your Arize AX space and select project **`vertexai-tracing-example`**. 2. You should see a new trace within \~30 seconds containing a `GenerateContent` LLM span with the prompt, response, and token usage attached. 3. If no traces appear, see [Troubleshooting](#troubleshooting). ## Troubleshooting * **No traces in Arize AX.** Confirm `ARIZE_SPACE_ID` and `ARIZE_API_KEY` are set in the same shell that runs `example.py`. Enable OpenTelemetry debug logs with `export OTEL_LOG_LEVEL=debug` and re-run. * **Vertex spans missing but other spans present.** `GoogleGenAIInstrumentor().instrument(...)` must run before any `from google import genai` import. Make sure `instrumentation.py` is the first import in your entry point. * **`401 UNAUTHENTICATED. API keys are not supported by this API.`** ADC isn't set up. Run `gcloud auth application-default login` (interactive) or set `GOOGLE_APPLICATION_CREDENTIALS` to a service account JSON. Vertex AI accepts only IAM-based auth, not API keys. * **`403 PERMISSION_DENIED. Permission denied on resource project ...`** The principal in your ADC doesn't have `roles/aiplatform.user` (or finer-grained Vertex permissions) on the project, or you authenticated with end-user credentials that have no quota project. Grant the role in the [IAM console](https://console.cloud.google.com/iam-admin/iam), then run `gcloud auth application-default set-quota-project `. * **`Your application has authenticated using end user credentials from Google Cloud SDK without a quota project`.** This warning means `gcloud auth application-default login` succeeded but you haven't set a quota project. Calls work but may hit `403` later. Fix with `gcloud auth application-default set-quota-project `. * **`404 NOT_FOUND` for the model.** The model isn't available in the region you set for `GOOGLE_CLOUD_LOCATION`. Check the [Vertex AI generative model availability matrix](https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models) and swap regions accordingly. * **Using the Gemini API instead of Vertex.** Drop `vertexai=True` and set `GEMINI_API_KEY` — see the [Google GenAI tracing](/ax/integrations/llm-providers/google-gen-ai/google-genai-tracing) doc for the API-key path. The same instrumentor captures both. * **Using the legacy `vertexai` SDK (`vertexai.GenerativeModel`).** That SDK is deprecated as of June 24, 2025 and will be removed June 24, 2026. Migrate to `google-genai` with `genai.Client(vertexai=True, ...)` as shown above. The OpenInference instrumentor for the legacy SDK (`openinference-instrumentation-vertexai`) still works against gapic but is no longer the recommended path. ## Resources