LLM
Represent language-model answer generation using retrieved context.
Category: Output. Registry label: LLM.
Purpose #
An LLM stage represents the generation component of a RAG pipeline. In a running system, it would receive a prompt and selected retrieval context and prepare a response.
Position in the graph #
Retriever → LLM → Answer / Output or Reranker → LLM → Answer / Output
Design guidance #
Use the configuration dialog to review the controls offered in the current editor. Consider prompt behavior and output needs, but do not assume the model has been contacted or that provider credentials are stored.
Current limitation #
No live text generation is performed by the current frontend pipeline.