# RAGLight

## Docs

- [Agentic RAG](https://docs.raglight.com/documentation/agentic-rag.md): Build Retrieval-Augmented Generation pipelines with reasoning and tool-aware agents.
- [AWS Bedrock](https://docs.raglight.com/documentation/bedrock.md): Use Claude, Titan, Llama and other Bedrock models for LLM inference and embeddings.
- [CLI](https://docs.raglight.com/documentation/cli.md): raglight chat, raglight agentic-chat, raglight serve — interactive RAG from your terminal.
- [Conversation History](https://docs.raglight.com/documentation/conversation-history.md): Multi-turn conversations with automatic history injection across all providers.
- [Embeddings](https://docs.raglight.com/documentation/embeddings.md): Configure embedding models and providers to index and retrieve your documents.
- [Hybrid Search](https://docs.raglight.com/documentation/hybrid-search.md): Combine semantic and keyword search with BM25 + Reciprocal Rank Fusion.
- [Knowledge Sources](https://docs.raglight.com/documentation/knowledge.md): How RAGLight discovers, loads, and prepares data for ingestion.
- [Observability with Langfuse](https://docs.raglight.com/documentation/langfuse.md): Trace every RAG call end-to-end — retrieve, rerank, and generate — directly in your Langfuse dashboard.
- [LLM Providers](https://docs.raglight.com/documentation/llm-providers.md): Configure and switch between different LLM backends in RAGLight.
- [RAG Pipelines](https://docs.raglight.com/documentation/rag.md): Build and run Retrieval-Augmented Generation pipelines in RAGLight.
- [Readers & Document Processors](https://docs.raglight.com/documentation/readers.md): How RAGLight reads, parses, and chunks files before embedding.
- [Query Reformulation](https://docs.raglight.com/documentation/reformulation.md): Automatically rewrite follow-up questions into standalone queries before retrieval.
- [REST API (raglight serve)](https://docs.raglight.com/documentation/rest-api.md): Deploy your RAG pipeline as a REST API without writing any code.
- [Settings](https://docs.raglight.com/documentation/settings.md): Global configuration, defaults, and provider identifiers in RAGLight.
- [Streaming](https://docs.raglight.com/documentation/streaming.md): Token-by-token output with generate_streaming() — drop-in alongside generate() on all providers.
- [Vector Stores](https://docs.raglight.com/documentation/vector-stores.md): How RAGLight stores embeddings and performs retrieval.
- [Agentic RAG & MCP](https://docs.raglight.com/examples/agentic.md): Give your RAG agent access to external tools via MCP.
- [Basic RAG](https://docs.raglight.com/examples/basic.md): Create a simple Q&A pipeline with local documents.
- [Multimodal RAG](https://docs.raglight.com/examples/multimodal.md): Ingest PDFs with images using Vision-Language Models.
- [Welcome to RAGLight](https://docs.raglight.com/index.md): Rapidly prototype, test, and run local-first Retrieval-Augmented Generation pipelines.
- [Getting Started](https://docs.raglight.com/overview/getting-started.md): Build your first RAG pipeline locally in a few minutes.
- [Introduction](https://docs.raglight.com/overview/introduction.md): A lightweight, local-first framework for rapidly prototyping and experimenting with Retrieval-Augmented Generation systems.

## OpenAPI Specs

- [openapi](https://docs.raglight.com/api-reference/openapi.json)