Weave Code
Ai Amazee Ai Platform Laravel Package
symfony/ai-amazee-ai-platform
Symfony AI bridge for the amazee.ai Platform. Connect Symfony AI to LiteLLM proxy endpoints and OpenAI-compatible providers through amazee.ai, enabling centralized AI access and management. Links to docs, issues, and contributions in the main Symfony AI repo.
amazee.ai LiteLLM platform bridge for Symfony AI
Frequently asked questions about Ai Amazee Ai Platform
- Can I use symfony/ai-amazeeai-platform directly in Laravel without Symfony’s AI bundle?
- No, this package is Symfony-first and requires its AI bundle. For Laravel, you’ll need to wrap Symfony’s `ClientInterface` in a facade or service provider. The core LiteLLM proxy functionality (multi-provider routing) can still work, but you’ll lose Symfony’s event system and DI integration.
- What Laravel versions support this package with custom integration?
- The package itself has no Laravel version constraints, but integration requires Laravel 8.50+ (for Symfony’s Psr-11 container support). Test thoroughly with your Laravel version, as Symfony’s `HttpClient` may need adjustments for Laravel’s HTTP stack.
- How do I configure amazee.ai keys and model routing in Laravel?
- Store your `AMAZEE_AI_KEY` in `.env` and create a `config/amazee_ai.php` file mirroring Symfony’s YAML structure. For routing, define fallbacks in the config (e.g., `gpt-4` → `mistral-7b`) and inject the client via Laravel’s service container.
- Will this package work with Laravel Livewire for real-time AI responses?
- Yes, but you’ll need to adapt Symfony’s `DeltaInterface` for Livewire’s streaming format. Use a custom event listener to convert DeltaInterface chunks into Livewire’s expected JSON structure. Test with small payloads first to avoid memory issues.
- Are there alternatives to this package for Laravel that avoid Symfony dependencies?
- For Laravel, consider `guzzlehttp/guzzle` with a custom proxy layer or packages like `spatie/laravel-ai` (if it supports LiteLLM). Direct Guzzle calls give you more control but lack built-in multi-provider routing and cost optimization.
- How does the LiteLLM proxy add latency compared to direct OpenAI API calls?
- Expect 50–200ms additional latency due to proxy routing. Benchmark with your workload—high-concurrency apps may need queue systems (e.g., Laravel Queues) or edge caching (e.g., Redis) to mitigate delays.
- Can I switch providers dynamically at runtime (e.g., OpenAI → Mistral) without code changes?
- Yes, the package’s routing layer supports runtime provider switching via configuration. Define fallbacks in `config/amazee_ai.php` (e.g., `fallback: mistral-7b`), and the proxy will handle the rest—no runtime logic required.
- What’s the maintenance risk of using this package in Laravel?
- High due to Symfony coupling. The package is actively developed (last release: 2026-06-16), but Laravel integrations may break with Symfony updates. Plan for forking or maintaining a Laravel-specific wrapper if long-term stability is critical.
- Does this package support Laravel’s queue system for async AI requests?
- Indirectly. Use Laravel Queues to defer HTTP calls to the LiteLLM proxy, but you’ll need to handle job failures and retries manually. Symfony’s AI bundle doesn’t natively integrate with Laravel Queues, so wrap the client in a job class.
- How do I handle errors if the amazee.ai proxy or LiteLLM fails?
- Implement a fallback mechanism in your Laravel service layer. Catch exceptions from the Symfony client and retry with a secondary provider (e.g., direct OpenAI API). Log errors to monitor proxy/LiteLLM availability.
Weaver
How can I help you explore Laravel packages today?