Skip to main content


Functionality related to Google Cloud Platform

Chat models

Gemini Models

Access Gemini models such as gemini-pro and gemini-pro-vision through the ChatGoogleGenerativeAI, or if using VertexAI, via the ChatVertexAI class.

npm install @langchain/google-genai

Configure your API key.

export GOOGLE_API_KEY=your-api-key
import { ChatGoogleGenerativeAI } from "@langchain/google-genai";

const model = new ChatGoogleGenerativeAI({
model: "gemini-pro",
maxOutputTokens: 2048,

// Batch and stream are also supported
const res = await model.invoke([
"What would be a good company name for a company that makes colorful socks?",

Gemini vision models support image inputs when providing a single human message. For example:

const visionModel = new ChatGoogleGenerativeAI({
model: "gemini-pro-vision",
maxOutputTokens: 2048,
const image = fs.readFileSync("./hotdog.jpg").toString("base64");
const input2 = [
new HumanMessage({
content: [
type: "text",
text: "Describe the following image.",
type: "image_url",
image_url: `data:image/png;base64,${image}`,

const res = await visionModel.invoke(input2);

Click here for the @langchain/google-genai specific integration docs

The value of image_url must be a base64 encoded image (e.g., data:image/png;base64,abcd124).

Vertex AI (Legacy)


See the legacy Google PaLM and VertexAI documentation here for chat, and here for LLMs.

Vector Store

Vertex AI Vector Search, formerly known as Vertex AI Matching Engine, provides the industry's leading high-scale low latency vector database. These vector databases are commonly referred to as vector similarity-matching or an approximate nearest neighbor (ANN) service.

import { MatchingEngine } from "langchain/vectorstores/googlevertexai";


  • Set up a Custom Search Engine, following these instructions
  • Get an API Key and Custom Search Engine ID from the previous step, and set them as environment variables GOOGLE_API_KEY and GOOGLE_CSE_ID respectively

There exists a GoogleCustomSearch utility which wraps this API. To import this utility:

import { GoogleCustomSearch } from "langchain/tools";

We can easily load this wrapper as a Tool (to use with an Agent). We can do this with:

const tools = [new GoogleCustomSearch({})];
// Pass this variable into your agent.

Help us out by providing feedback on this documentation page: