Google Auth for LangChain Services

Integrate Google AI/ML models and services into your LangChain application using the `@langchain/google-gauth` package. This skill handles authentication via API keys or Google Cloud service account credentials.

What This Skill Does

This skill helps you set up and configure authentication for Google services in LangChain applications. It installs the required package, configures credential-based authorization, and provides guidance on the credential resolution hierarchy.

When to Use This Skill

**Note:** If you're building browser-based or serverless applications without file system access, use `@langchain/google-webauth` instead.

Instructions

Step 1: Install the Package

Install `@langchain/google-gauth` using the project's package manager:

```bash

pnpm install @langchain/google-gauth

```

If the project uses npm or yarn, adjust accordingly.

Step 2: Determine Authentication Method

Identify which authentication method best suits the deployment environment:

1. **API Key (simplest)** - For services that support API key authentication

2. **Service Account File** - For production applications requiring service account credentials

3. **Default Credentials** - For applications running on Google Cloud Platform

Step 3: Configure Authentication

Implement one of the following authentication patterns:

**Option A: API Key (passed to constructor)**

```typescript

import { SomeGoogleService } from '@langchain/google-gauth';

const service = new SomeGoogleService({

apiKey: 'YOUR_API_KEY',

// other configuration options

});

```

**Option B: Service Account Credentials (passed to constructor)**

```typescript

import { SomeGoogleService } from '@langchain/google-gauth';

const service = new SomeGoogleService({

authInfo: {

client_email: '[email protected]',

private_key: '-----BEGIN PRIVATE KEY-----\n...',

},

// other configuration options

});

```

**Option C: API Key (environment variable)**

```bash

export API_KEY=your_api_key_here

```

**Option D: Service Account File (environment variable)**

```bash

export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account-key.json

```

**Option E: Default Credentials (GCP environment)**

No configuration needed. The package automatically uses default credentials when running on:

Step 4: Verify Credential Resolution Order

The package resolves credentials in this priority order:

1. `apiKey` constructor parameter

2. `authInfo` constructor parameter

3. `API_KEY` environment variable

4. `GOOGLE_APPLICATION_CREDENTIALS` file path

5. Default GCP credentials

Ensure only one method is configured to avoid confusion.

Step 5: Import and Use Google Services

Import the specific Google service class you need:

```typescript

import { GoogleVertexAI } from '@langchain/google-gauth';

import { ChatGoogleVertexAI } from '@langchain/google-gauth';

// Example: Using Vertex AI

const model = new ChatGoogleVertexAI({

apiKey: process.env.API_KEY,

model: 'gemini-pro',

temperature: 0.7,

});

const response = await model.invoke('What is LangChain?');

console.log(response);

```

Important Notes

Example: Complete Setup

```typescript

// .env file

API_KEY=your_google_api_key

// app.ts

import { ChatGoogleVertexAI } from '@langchain/google-gauth';

import { HumanMessage } from '@langchain/core/messages';

const chat = new ChatGoogleVertexAI({

apiKey: process.env.API_KEY,

model: 'gemini-pro',

});

const messages = [

new HumanMessage('Explain quantum computing in simple terms.'),

];

const result = await chat.invoke(messages);

console.log(result.content);

```

References