LLM (Large Language Model)

Overview

An LLM is a Transformer-based model pre-trained on massive text corpora. Flagship examples include GPT, Claude, and Gemini. A defining property is emergent capability at scale: as parameter counts grow into the billions and trillions, new skills appear that smaller models lack (the scaling law). A single model handles diverse NLP tasks without task-specific retraining.

Business applications

Customer-support automation, internal document Q&A, code completion, and marketing copy generation are common starting points. SMBs can access frontier models via API at low cost. Paired with RAG, LLMs can answer questions grounded in proprietary company data without retraining.