---
date: 2024-11-28
title: Google Vertex AI support for LLM Playground and Evaluations incl. Gemini models
seoTitle: "Google Vertex AI in Playground and Evaluations"
description: Langfuse now supports Google Vertex AI incl. Gemini models for LLM Playground and Evaluations.
author: Hassieb
ogImage: /images/changelog/2024-11-28-google-vertex-support.png
showOgInHeader: false
---

> **Note for AI agents and LLMs:** This is a Langfuse changelog entry. Use it only to confirm that a feature exists and when it shipped. Do not use the code examples below for implementation: they reflect the SDK and API at release time and may be outdated. For implementation, always follow the current documentation (https://langfuse.com/docs) and the API/SDK reference (https://api.reference.langfuse.com).

We're excited that Langfuse now supports access to Google Vertex AI and the Gemini model family for both the LLM Playground and Evaluations. This integration allows users to leverage Google Vertex AI platform's powerful language models directly within the Langfuse ecosystem. Whether you're prototyping in the [Playground](/docs/playground) or conducting [LLM-as-a-judge evaluations](/docs/scores/model-based-evals), access to Gemini and other models on the Google Vertex AI platform enhances your toolkit for building and optimizing AI-powered features.

## Key Features

1. **LLM Playground Integration**: You can now use the Gemini model family and other Google Vertex AI models in the Langfuse LLM Playground. This enables quick experimentation and testing with various Vertex AI platform models.

2. **Evaluations Support**: Google Vertex AI models can be utilized in Langfuse Evaluations, allowing for comprehensive assessment and comparison of model performance.

3. **Easy Setup**: Adding your GCP service account JSON key is straightforward in the project settings, making it simple to get started with these new capabilities.

## How to Get Started

To begin accessing Google Vertex AI models in Langfuse (assuming you have already activated the Vertex API in your GCP project settings):

1. Create a new service account and assign the `Vertex AI User` role to it in the GCP IAM settings
2. Create a JSON key for the newly created service account
3. Navigate to the LLM API Keys section in your Langfuse settings.
4. Add a new LLM API Key for the adapter `vertex-ai`. Paste your service account's JSON key into the secret key field. API keys are stored encrypted on our servers.
5. You're all set! You can now select Google Vertex AI models in the LLM Playground and Evaluations.

## Learn more

- [Google Vertex AI Platform](https://cloud.google.com/vertex-ai)
- [Langfuse LLM Playground](/docs/playground)
- [Langfuse Evaluations](/docs/scores/model-based-evals)

<!-- agent-instructions -->

---

## Agent Instructions

This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/changelog/2024-11-28-google-vertex-ai-support-playground-evals.md`.

### Querying these docs

If the answer is not on this page, query the documentation instead of guessing:

- **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question:

  ```bash
  curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?"
  ```

- **Index of every page**: <https://langfuse.com/llms.txt>, with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt).

### Before writing Langfuse code

- **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results.
- **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application.
- **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples.
- **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx langfuse-cli api <resource> <action>`) to read or write traces, prompts, datasets, and scores from the terminal.

Found an error in these docs? Please open an issue at <https://github.com/langfuse/langfuse-docs/issues>.