> ## Documentation Index
> Fetch the complete documentation index at: https://docs.raglight.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Welcome to RAGLight

> Rapidly prototype, test, and run local-first Retrieval-Augmented Generation pipelines.

export const Hero = () => <div className="flex flex-col items-center justify-center text-center py-16 px-4">
    <h1 className="text-4xl md:text-6xl font-extrabold tracking-tight text-slate-900 dark:text-slate-100 mb-6">
      Build and experiment with <span style={{
  color: '#FF4017'
}}>RAG</span>,
      locally.
    </h1>
    <p className="text-lg md:text-xl text-slate-600 dark:text-slate-400 max-w-2xl mb-10 leading-relaxed">
      RAGLight is a lightweight, modular Python framework for rapidly
      prototyping and running Retrieval-Augmented Generation systems.
    </p>
    <div className="flex gap-4">
      <a href="/introduction" className="px-6 py-3 rounded-full font-semibold text-white transition-all hover:opacity-90" style={{
  backgroundColor: '#FF4017'
}}>
        Get Started
      </a>
      <a href="https://github.com/Bessouat40/RAGLight" target="_blank" className="px-6 py-3 rounded-full font-semibold border border-slate-300 dark:border-slate-700 hover:bg-slate-50 dark:hover:bg-slate-800 transition-all text-slate-700 dark:text-slate-200">
        View on GitHub
      </a>
    </div>
  </div>;

<Hero />

<div className="my-12">
  <div className="text-center mb-6">
    <p className="text-xs font-bold uppercase tracking-widest text-slate-500">
      Quick Start Examples
    </p>
  </div>

  {/* Wrapper pour centrer et limiter la largeur du code */}

  <div className="max-w-3xl mx-auto text-left">
    <CodeGroup>
      ```bash CLI Quickstart theme={null}
      # Install RAGLight
      pip install raglight

      # Launch interactive RAG setup
      raglight chat

      # Optional Agentic CLI
      raglight agentic-chat

      # Deploy as a REST API (no code required)
      raglight serve
      # Deploy a streamlit application
      raglight serve --ui
      ```

      ```python Python Minimal RAG theme={null}
      from raglight.rag.simple_rag_api import RAGPipeline
      from raglight.config.settings import Settings
      from raglight.config.rag_config import RAGConfig

      # 1. Setup
      Settings.setup_logging()

      # 2. Configure with Defaults
      config = RAGConfig(
        provider=Settings.OLLAMA,
        llm="llama3",
        knowledge_base="./data"
      )

      # 3. Build & Run
      pipeline = RAGPipeline(config)
      pipeline.build()

      response = pipeline.generate("Explain RAGLight basics.")
      print(response)
      ```
    </CodeGroup>
  </div>
</div>

<div className="py-10">
  {/* Wrapper pour centrer les cartes et passer à 2 colonnes */}

  <div className="max-w-4xl mx-auto">
    <CardGroup cols={2}>
      <Card title="Local-First" icon="laptop" iconType="duotone" color="#FF4017">
        Build and iterate without dependency on remote services. Privacy by
        design.
      </Card>

      <Card title="Modular Pipelines" icon="puzzle-piece" iconType="duotone" color="#FF4017">
        Swap embeddings, vector stores, and LLMs freely (Ollama, Chroma, etc).
      </Card>
    </CardGroup>
  </div>
</div>

<div className="text-center text-sm text-slate-500 mt-16 pb-10">
  Open-source, local-first, and ready for experimentation.
</div>