Vertex AI

You can also use Google Cloud's Vertex AI models, including Gemini or your own fine-tuned models, with DeepEval.

Command Line

Run the following command in your terminal to configure your deepeval environment to use Gemini models through Vertex AI for all metrics.

deepeval set-gemini \
    --model=<model> \ # e.g. "gemini-2.5-flash"
    --project=<project_id> \
    --location=<location> # e.g. "us-central1"

Python

Alternatively, you can specify your model directly in code using GeminiModel from DeepEval's model collection. By default, model is set to gemini-2.5-pro.

from deepeval.models import GeminiModel
from deepeval.metrics import AnswerRelevancyMetric

model = GeminiModel(
    model="gemini-2.5-pro",
    project="Your Project ID",
    location="us-central1",
    temperature=0
)

answer_relevancy = AnswerRelevancyMetric(model=model)

There are ZERO mandatory and SEVEN optional parameters when creating an GeminiModel through Vertex AI:

• [Optional] model: A string specifying the name of the Gemini model to use. Defaults to GEMINI_MODEL_NAME if not passed; raises an error at runtime if unset.
• [Optional] temperature: A float specifying the model temperature. Defaults to TEMPERATURE if not passed; falls back to 0.0 if unset.
• [Optional] project: A string specifying the Google Cloud project ID for Vertex AI. Defaults to GOOGLE_CLOUD_PROJECT if not passed.
• [Optional] location: A string specifying the Google Cloud location for Vertex AI. Defaults to GOOGLE_CLOUD_LOCATION if not passed.
• [Optional] service_account_key: A JSON string containing the service account key for authentication when using Vertex AI. This string can be either the path to a service account key file or the raw JSON string. Defaults to GOOGLE_SERVICE_ACCOUNT_KEY if not passed.
• [Optional] use_vertexai: A boolean to explicitly force Vertex AI (True) or Gemini API-key mode (False); if not passed, defaults to GOOGLE_GENAI_USE_VERTEXAI and otherwise falls back to auto-detection via project and location.
• [Optional] generation_kwargs: A dictionary of additional generation parameters supported by your model provider.

Available Vertex AI Models

Below is a list of commonly used Gemini models:

gemini-3-pro-preview gemini-3-flash-preview gemini-2.5-pro gemini-2.5-flash gemini-2.5-flash-lite gemini-2.0-flash gemini-2.0-flash-lite gemini-pro-latest gemini-flash-latest gemini-flash-lite-latest