RAG Chat

TrustRAG's RAG (Retrieval-Augmented Generation) chat lets you ask questions about your uploaded documents and receive grounded answers with citations.

How It Works

1. You ask a question in a workspace conversation
2. Hybrid retrieval searches your documents using vector similarity + full-text search
3. Relevant chunks are retrieved and ranked using Reciprocal Rank Fusion (RRF)
4. The LLM generates an answer grounded in the retrieved chunks
5. Citations are extracted and stored, linking each claim to its source

Supported LLM Providers

TrustRAG supports any OpenAI-compatible API:

• OpenAI (GPT-4, GPT-3.5)
• Ollama (local models)
• Anthropic
• Custom endpoints

Configure providers in Settings → Model Configuration.

Streaming Responses

Chat responses stream in real-time via Server-Sent Events (SSE). Citations are processed and stored after the response completes.