Weave Code
Ai Bundle Laravel Package
lingoda/ai-bundle
Lingoda AI Bundle is a Symfony package that makes it easy to add AI-powered features to PHP applications. It provides configurable services and integration points for building intelligent workflows with minimal setup. Ideal for teams experimenting with AI in existing Symfony projects.
Symfony Bundle for Lingoda AI PHP SDK with full framework integration
Frequently asked questions about Ai Bundle
- Can I use lingoda/ai-bundle in a Laravel application?
- You can install it, but it will not do anything useful out of the box. It is a Symfony bundle whose value is framework glue: a DependencyInjection extension, a configuration tree, and container-compiled autowired services, none of which Laravel's container or bootstrap lifecycle will read. The recommended path for Laravel is to consume the underlying Lingoda AI PHP SDK directly and bind it in your own service provider. Pulling in symfony/framework-bundle just to satisfy a bundle reverses the dependency direction and adds weight for no runtime benefit.
- Is lingoda/ai-bundle a Laravel package or a Symfony bundle?
- It is a Symfony bundle, published as lingoda/ai-bundle and intended to be registered in bundles.php inside a Symfony kernel. Its advertised features, such as dependency injection, autowiring, and configuration-driven model and API key management, are Symfony container concepts. Laravel already provides its own equivalents, so treat the package as a reference implementation rather than a drop-in dependency.
- How do I install and register lingoda/ai-bundle in a Symfony project?
- The typical flow is composer require lingoda/ai-bundle, register the bundle in config/bundles.php, and create a config/packages/lingoda_ai.yaml file for your provider, model, and credentials. The bundle then exposes configured client services you can inject into your own services. If you are not running Symfony, this installation path does not apply to you.
- What is the best way to add Lingoda AI to Laravel without the Symfony bundle?
- Write a small Laravel service provider that binds the SDK client as a singleton, and mirror the bundle's configuration tree in a config/ai.php file fed by environment variables. If the SDK is PSR-18 and PSR-17 based and stateless, that alone gets you most of the parity in well under a day. Anything the bundle registers beyond a configured client, such as HTTP middleware for retries, logging, or telemetry, should be ported explicitly rather than assumed. Budget roughly two to five days if you need streaming, retries, and test coverage.
- Which PHP and Laravel versions does lingoda/ai-bundle support?
- The bundle does not target Laravel versions at all, so there is no Laravel support matrix to check. You need to read its composer.json for the PHP floor and the supported Symfony major versions, and remember that Laravel only ships symfony/console, http-foundation, finder, and process transitively, not framework-bundle or dependency-injection. Because the release cadence tracks a single vendor's internal roadmap, verify the constraints on Packagist before committing to it in a long-lived application.
- How should I store AI API keys and model settings in a Laravel app?
- Keep credentials in environment variables and reference them from config/ai.php using env() calls, never hard-coded in service classes. Cache the resolved configuration in production with php artisan config:cache so the env lookups are not repeated per request. If you support multiple tenants or providers, model that as a keyed array in the config file and resolve the right profile in your service provider rather than scattering conditionals through your code.
- Does lingoda/ai-bundle support streaming responses for chat interfaces?
- Streaming support is not documented in the package README, so you must confirm it against the underlying SDK before designing a chat UX around it. If the SDK exposes SSE or token streaming, you can bridge it to Laravel's StreamedResponse, but that bridging code is yours to write and test. Treat streaming as an open question during evaluation rather than a feature you can assume.
- How do I test an AI integration in Laravel without calling the real API?
- Bind a fake implementation of your client interface in the container during tests, or swap in a mock PSR-18 client that returns canned responses. That keeps your test suite fast, deterministic, and free of network flakiness or token costs. Keep a small number of optional contract tests against the real endpoint, gated behind an environment flag, to catch upstream response shape changes.
- What are the alternatives to lingoda/ai-bundle for AI features in Laravel?
- Most Laravel teams either use a Laravel-native AI package with an OpenAI-compatible client or write a thin provider-agnostic abstraction over PSR-18 themselves. If you specifically want Lingoda's platform, the cleanest option is the Lingoda AI PHP SDK plus your own service provider. The bundle is best treated as a reference for auth, config defaults retry policy, and service composition rather than as the integration unit.
- Is lingoda/ai-bundle actively maintained and safe to use in production?
- It is a single-vendor artifact with roughly one star and no known dependents, so no third party has validated it in production outside the vendor's own estate. The README is effectively the only documentation, the MIT license provides no warranty or support obligation, and there is product-level lock-in to the Lingoda AI platform for endpoints, auth, pricing, and rate limits. Its listed release date also looks anomalous and should be verified against Packagist before you rely on it in planning.
Weaver
How can I help you explore Laravel packages today?