LLPhant

Why we included this project

PHP developers often hit a wall when they want to add generative AI to an existing Symfony or Laravel app, since most of the ecosystem assumes Python. LLPhant fills that gap with a Composer-installable library that handles chat completions, embeddings, and vector retrieval behind a simple API, and it works with OpenAI, Anthropic, Mistral, Ollama, and other OpenAI-compatible services. Provider swaps happen through configuration rather than code rewrites, which keeps things flexible if you start with one model and move to another. The framework also covers agents and structured output, the building blocks that usually take real effort to wire together by hand. For a PHP team that wants LLM features in production without stitching together several SDKs, this is a practical, coherent place to start.