execution-gemini

Google Gemini provider implementation for LLM execution. Implements the Provider interface from the execution package.

Installation

npm install execution-gemini @google/generative-ai

Usage

import { GeminiProvider, createGeminiProvider } from 'execution-gemini';

// Create provider
const provider = createGeminiProvider();

// Execute a request
const response = await provider.execute(
  {
    model: 'gemini-1.5-pro',
    messages: [
      { role: 'system', content: 'You are helpful.' },
      { role: 'user', content: 'Hello!' }
    ],
    addMessage: () => {},
  },
  {
    apiKey: process.env.GEMINI_API_KEY,
    temperature: 0.7,
  }
);

console.log(response.content);
console.log(response.usage); // { inputTokens: X, outputTokens: Y }

Supported Models

The provider supports all Gemini models:

• Gemini 1.5 Pro
• Gemini 1.5 Flash
• Gemini 1.0 Pro

API Key

Set via:

1. options.apiKey parameter
2. GEMINI_API_KEY environment variable

Features

• System instruction support
• Multi-turn conversation handling
• Structured output via JSON schema
• Token usage tracking

Response Format

interface ProviderResponse {
  content: string;
  model: string;
  usage?: {
    inputTokens: number;
    outputTokens: number;
  };
}

Related Packages

• execution - Core interfaces (no SDK dependencies)
• execution-openai - OpenAI provider
• execution-anthropic - Anthropic provider

License

Apache-2.0