LLM (Large Language Model)

Large Language Models (LLMs) are AI systems trained on enormous text datasets using transformer neural network architecture. They learn the statistical patterns of language, enabling them to generate coherent, contextually appropriate text responses.

Major LLMs include: GPT-4 (OpenAI), Claude (Anthropic), Gemini (Google), Llama (Meta, open-source), and Mistral. These models are distinguished by their parameter count (7B to 1T+), training data, and fine-tuning approaches.

LLMs have capabilities beyond text generation: reasoning, code generation, mathematical problem-solving, translation, summarization, and multimodal understanding (processing images and audio alongside text). They are the foundation of the generative AI revolution.