Gemini Dart

A comprehensive Dart package for integrating Google's Gemini AI models with support for text, image, and video processing.

Features

• 🤖 Multi-modal AI Integration: Support for text, image, and video inputs
• 🔧 Easy Integration: Simple API for Dart applications
• 🔄 Streaming Responses: Real-time text generation with streaming support
• 🛡️ Type Safety: Strongly typed models and responses
• 🔧 Configurable: Extensive configuration options for different use cases
• 📦 Cross-platform: Works on iOS, Android, Web, and Desktop
• 🚀 Performance Optimized: Efficient file handling and caching mechanisms

Installation

Add this package to your pubspec.yaml:

dependencies:
  gemini_dart: ^0.1.0

Then run:

dart pub get

Quick Start

Basic Text Generation

import 'package:gemini_dart/gemini_dart.dart';

void main() async {
  // Initialize the client
  final client = GeminiClient();
  await client.initialize('YOUR_API_KEY');

  // Generate text
  final response = await client.generateText(prompt: 'Hello, Gemini!');
  print(response.text);

  // Clean up
  client.dispose();
}

Multi-modal Content

import 'package:gemini_dart/gemini_dart.dart';

void main() async {
  final client = GeminiClient();
  await client.initialize('YOUR_API_KEY');

  // Combine text and image
  // Analyze an image with optional prompt
  final response = await client.imageHandler.analyzeImage(
    imageData: imageBytes,
    mimeType: 'image/jpeg',
  );
  print(response.text);
}

Advanced Usage

import 'package:gemini_dart/gemini_dart.dart';

void main() async {
  final client = GeminiClient();
  await client.initialize('YOUR_API_KEY');

  // Advanced configuration
  final config = GenerationConfig(
    temperature: 0.7,
    maxOutputTokens: 1000,
    topP: 0.9,
  );

  final response = await client.generateText(
    prompt: 'Write a creative story',
    config: config,
  );

  print(response.text);
  client.dispose();
}

Configuration

Basic Configuration

final config = GeminiConfig(
  timeout: Duration(seconds: 30),
  maxRetries: 3,
  enableLogging: true,
);

final client = GeminiClient();
await client.initialize('YOUR_API_KEY', config: config);

Generation Parameters

final generationConfig = GenerationConfig(
  temperature: 0.7,
  maxOutputTokens: 1000,
  topP: 0.9,
  topK: 40,
);

final response = await client.generateText(
  prompt: 'Write a story about AI',
  config: generationConfig,
);

Supported Content Types

Text

• Simple text prompts
• Multi-turn conversations
• Streaming responses

Images

• JPEG, PNG, WebP formats
• Automatic resizing for size limits
• Multi-image analysis

Video

• MP4, MOV formats
• Large file upload with progress tracking
• Frame-by-frame analysis

Error Handling

The package provides comprehensive error handling:

try {
  final response = await client.generateText(prompt: 'Hello');
} on GeminiAuthException catch (e) {
  print('Authentication error: ${e.message}');
} on GeminiRateLimitException catch (e) {
  print('Rate limit exceeded. Retry after: ${e.retryAfter}');
} on GeminiException catch (e) {
  print('Gemini error: ${e.message}');
}

Advanced Features

Streaming Responses

Get real-time text generation with streaming support.

File Processing

Handle images and videos with automatic format detection and optimization.

Conversation Context

Maintain conversation history for multi-turn interactions.

Examples

Check out the /example directory for complete examples:

• Basic text generation
• Multi-modal content processing
• Image generation and analysis
• Video analysis workflow

API Reference

For detailed API documentation, visit pub.dev documentation.

Requirements

• Dart SDK: >=3.0.0 <4.0.0
• Google AI API key

Getting an API Key

1. Visit Google AI Studio
2. Create a new API key
3. Add the key to your environment or configuration

Contributing

We welcome contributions! Please see our Contributing Guide for details.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

• 📖 Documentation
• 🐛 Issue Tracker
• 💬 Discussions

Changelog

See CHANGELOG.md for a list of changes in each version.