Skip to main content

Google LLM Adapter for RTP Agent

Google provides access to the Gemini family of models (Pro and Flash), which are natively multimodal and optimized for high-speed, low-cost intelligence.

Installation

Add the adapter to your Go project:

go get github.com/cavos-io/rtp-agent/adapter/google

Authentication

Set the required environment variables in your .env file. Refer to the Google developer documentation for acquiring the necessary API keys and tokens.

GOOGLE_API_KEY=your_api_key_here

Usage

Below is a basic conceptual example demonstrating how to initialize the Google LLM adapter within an RTP Agent session:

package main

import (
"context"
"log"
"os"

"github.com/cavos-io/rtp-agent/adapter/google"
"github.com/cavos-io/rtp-agent/core/agent"
)

func main() {
ctx := context.Background()

// Initialize the Google LLM adapter
llmProvider, err := google.NewProvider(
os.Getenv("GOOGLE_API_KEY"),
)
if err != nil {
log.Fatalf("failed to initialize google adapter: %v", err)
}

// Create and configure the RTP agent session
session := agent.NewSession(
agent.WithLLM(llmProvider),
)

// Start the session
if err := session.Start(ctx); err != nil {
log.Fatalf("session failed: %v", err)
}
}