LLM (Large Language Model)

An LLM (large language model) is an AI model trained on vast amounts of text to understand and generate language. Simplified, it predicts the most likely next word at each step — and can thereby answer questions, summarise texts, translate or draft. Examples are the models behind assistants such as ChatGPT, Claude or Mistral.

Why LLMs matter

LLMs are the engine of modern conversational AI. They make it possible to interact with software in natural language instead of filling in forms. Their weakness: an LLM only “knows” what was in its training data, and tends to fill knowledge gaps with convincing-sounding inventions — so-called hallucinations. That’s why, for reliable answers, LLMs are combined with RAG.

LLMs at Kyros

Kyros doesn’t lock you into a single model: a model router gives access to several providers, with transparent per-message credit billing and automatic fallback if a provider goes down. With Compare Models you put two configurations side by side live and send the same question to both — so you pick the model that best fits your use case.