Weave Code
Client Laravel Package
openai-php/client
Community-maintained PHP client for the OpenAI API. Install via Composer and interact with models, responses, chat, images, audio, files, and more with a clean, typed interface—ideal for Laravel and modern PHP apps.
⚡️ OpenAI PHP is a supercharged community-maintained PHP API client that allows you to interact with OpenAI API.
Frequently asked questions about Client
- How do I install OpenAI PHP in a Laravel project?
- Use Composer to install the package with `composer require openai-php/client`. Laravel’s built-in HTTP client (PSR-18 compliant) works out of the box, so no additional setup is required unless you’re using a custom HTTP client like Guzzle. Register the client in Laravel’s service container via `config/app.php` or a service provider for dependency injection.
- What Laravel versions does OpenAI PHP support?
- The package is designed for PHP 8.2+ and integrates seamlessly with Laravel 9.x and 10.x. It leverages PSR-18 HTTP clients, which Laravel’s built-in `Http` facade already supports. For older Laravel versions, ensure your HTTP client is PSR-18 compliant or use a compatible adapter like `php-http/guzzle7-adapter`.
- Can I use OpenAI PHP for streaming responses like chat completions in real-time?
- Yes, the package supports streaming responses via methods like `createStreamed()`. However, Laravel’s middleware may interfere with streaming contexts. Use Laravel’s `streamDownload()` or create a custom controller to handle streaming responses without middleware blocking. Test thoroughly in a staging environment to ensure compatibility.
- How should I manage OpenAI API keys securely in Laravel?
- Store API keys in Laravel’s `.env` file (e.g., `OPENAI_API_KEY`) and access them via `config('services.openai.key')`. For multi-key setups (e.g., failover or cost optimization), use Laravel’s `config()` with environment-specific keys or a package like `spatie/laravel-env-editor` to manage multiple keys dynamically. Avoid hardcoding keys in your application.
- Does OpenAI PHP support rate limiting and token usage tracking?
- The package itself doesn’t enforce rate limits, but you can implement middleware or Laravel tasks to monitor `usage->totalTokens` and trigger alerts when thresholds are exceeded. Use Laravel’s queue system to batch requests and avoid hitting rate limits. For production, consider integrating with OpenAI’s usage logs or a third-party monitoring tool.
- How do I handle deprecated OpenAI API resources (e.g., assistants) in Laravel?
- Use abstract classes or interfaces to isolate API changes. For example, create an `OpenAIAssistantsInterface` that your Laravel service implements. When OpenAI deprecates a resource, update the interface and adapter layer without breaking existing Laravel logic. This approach also simplifies testing and future migrations.
- Can I cache OpenAI API responses in Laravel (e.g., model listings) for performance?
- Yes, cache frequent responses like model listings using Laravel’s cache drivers (e.g., Redis or file cache). Wrap API calls in a service class and use `Cache::remember()` to store responses. Set appropriate TTL values based on OpenAI’s API update frequency. For real-time data (e.g., chat responses), avoid caching unless absolutely necessary.
- How do I test OpenAI PHP in Laravel with mocked HTTP responses?
- Use Laravel’s HTTP client mocking (e.g., `Http::fake()`) or libraries like `Mockery` or `Vcr` to record and replay API responses. For unit tests, inject a mocked HTTP client into your service container. Validate responses using Laravel’s `assertJson()` or PHPUnit assertions. This ensures your Laravel logic works independently of OpenAI’s API.
- What are the alternatives to OpenAI PHP for Laravel, and when should I consider them?
- Alternatives include the official OpenAI PHP SDK (`openai`) or custom implementations using Guzzle directly. Choose OpenAI PHP for its Laravel-friendly design, typed interfaces, and community support. Use the official SDK if you need tighter alignment with OpenAI’s latest features or prefer minimal dependencies. For legacy projects, a custom Guzzle wrapper may suffice if you’re already using it.
- How do I handle long-running OpenAI API requests (e.g., file uploads, fine-tuning) in Laravel?
- Offload long-running tasks to Laravel queues using `dispatch()` with `ShouldQueue` interfaces. For file uploads, use `Storage::disk()` to handle large files before sending them to OpenAI. Monitor queue jobs with Laravel Horizon or `queue:work` to ensure reliability. Consider async processing for fine-tuning jobs to avoid timeouts in web requests.
Weaver
How can I help you explore Laravel packages today?