README

A simple and effective integration between Laravel and ChatGPT. This package provides a clean, developer-friendly interface for interacting with the OpenAI API, including support for chat, text completions, and image analysis with GPT-4 Vision.

Features

✅ Chat Completions – Conversational AI with memory
✅ Text Completions – Continue or generate texts
✅ Vision Analysis – Image understanding with GPT-4 Vision
✅ Rate Limiting – Configurable request throttling
✅ Retry Logic – Automatic retries on failures
✅ Multiple Models – GPT-3.5, GPT-4, and GPT-4 Vision supported
✅ Laravel Integration – Includes Service Provider and Facade
✅ Auto Discovery – Automatically registered in Laravel

Installation

Install the package via Composer:

composer require sysborg/chatgpt

Publish Configuration

To publish the config file:

php artisan vendor:publish --tag=chatgpt-config

Set Your API Key

Add your OpenAI API key in the .env file:

OPENAI_API_KEY=your_openai_api_key_here

Configuration

Customize config/chatgpt.php to match your needs:

return [
    'api_key' => env('OPENAI_API_KEY'),
    'base_url' => env('OPENAI_BASE_URL', 'https://api.openai.com/v1'),
    'default_model' => env('OPENAI_DEFAULT_MODEL', 'gpt-3.5-turbo'),
    'timeout' => env('OPENAI_TIMEOUT', 60),
    'max_tokens' => env('OPENAI_MAX_TOKENS', 1000),
    'temperature' => env('OPENAI_TEMPERATURE', 0.7),
    'retry_attempts' => env('OPENAI_RETRY_ATTEMPTS', 2),
    'retry_delay' => env('OPENAI_RETRY_DELAY', 1),
    'rate_limit' => env('OPENAI_RATE_LIMIT', 60), // 0 = unlimited
];

Basic Usage

Simple Chat

use Sysborg\ChatGPT\Facades\ChatGPT;

$response = ChatGPT::chat('Hello, how are you?');
echo $response->getContent();

Chat With History

$messages = [
    ['role' => 'user', 'content' => 'What is the capital of Brazil?'],
    ['role' => 'assistant', 'content' => 'The capital of Brazil is Brasília.'],
    ['role' => 'user', 'content' => 'What is the population?']
];

$response = ChatGPT::chatWithHistory($messages);
echo $response->getContent();

Text Completion

$response = ChatGPT::completion('Once upon a time in a distant kingdom');
echo $response->getContent();

Image Analysis

$response = ChatGPT::vision(
    'https://example.com/image.jpg',
    'Describe this image in detail'
);

// With base64 image
$base64 = base64_encode(file_get_contents('/path/to/image.jpg'));
$response = ChatGPT::visionFromBase64(
    $base64,
    'What do you see in this image?'
);

echo $response->getAnalysis();

Dynamic Configuration

$response = ChatGPT::setModel('gpt-4')
    ->setTemperature(0.9)
    ->setMaxTokens(2000)
    ->chat('Tell a creative story');

Handling Responses

Chat Response

$response = ChatGPT::chat('Hi!');

echo $response->getContent();
echo $response->getModel();
echo $response->getTotalTokens();
echo $response->getFinishReason();

if ($response->isComplete()) {
    echo "Completed successfully";
}

if ($response->isTruncated()) {
    echo "Response was truncated";
}

Vision Response

$response = ChatGPT::vision($imageUrl, 'Analyze this image');

echo $response->getAnalysis();
print_r($response->getDetectedEntities());
echo $response->getSummary();

if ($response->hasSafetyConcerns()) {
    echo "This image may contain sensitive content";
}

Error Handling

use Sysborg\ChatGPT\Exceptions\ChatGPTException;
use Sysborg\ChatGPT\Exceptions\RateLimitException;

try {
    $response = ChatGPT::chat('Hi!');
} catch (RateLimitException $e) {
    echo "Rate limit exceeded. Retry in: " . $e->getRetryAfter() . " seconds";
} catch (ChatGPTException $e) {
    echo "API error: " . $e->getMessage();
    print_r($e->getContext());
}

Rate Limiting

Built-in rate limit handling:

$status = ChatGPT::getRateLimitStatus();
echo "Requests remaining: " . $status['remaining'];

ChatGPT::resetRateLimit(); // Useful for testing

Available Models

$models = ChatGPT::getAvailableModels();
print_r($models);

Example Output:

[
    'gpt-3.5-turbo',
    'gpt-3.5-turbo-16k',
    'gpt-4',
    'gpt-4-turbo',
    'gpt-4-vision-preview',
    'gpt-4-32k'
]

Advanced Examples

Chat with Custom Config

$response = ChatGPT::chatWithHistory([
    ['role' => 'system', 'content' => 'You are a programming assistant.'],
    ['role' => 'user', 'content' => 'How do I create a REST API in Laravel?']
], [
    'model' => 'gpt-4',
    'temperature' => 0.3,
    'max_tokens' => 1500,
    'top_p' => 0.9
]);

Vision Analysis with Settings

$response = ChatGPT::vision($imageUrl, 'Analyze this image', [
    'model' => 'gpt-4-vision-preview',
    'detail' => 'high', // Options: low, high, auto
    'max_tokens' => 1000
]);

Contributing

Contributions are welcome! Please open an issue or submit a pull request.

License

This package is licensed under the MIT License.

Author

Anderson Arruda 📧 andmarruda@gmail.com 🐙 @andmarruda on GitHub

Support

If you encounter any issues or have questions, please open a GitHub Issue.

sysborg / chatgpt

Maintainers

Details