Glossary
LLM (Large Language Model)
A type of AI model trained on massive amounts of text data, capable of understanding and generating natural language at human-comparable quality for most tasks.
An LLM (Large Language Model) is a neural network — usually a transformer architecture — trained on hundreds of billions to trillions of tokens of text data. The result is a model that can understand and generate natural language across most domains, follow complex instructions, and reason about novel situations.
Current production-grade LLMs as of 2026:
- Claude (Anthropic): Opus 4.7, Sonnet 4.6, Haiku 4.5 — the default choice for serious automation work
- GPT-4o / GPT-5 (OpenAI): widely deployed, strong general capability
- Gemini 2 Pro / Flash (Google): strong long-context, native multimodal
- Llama 4 (Meta): open-weights option for self-hosting
Key properties relevant to automation:
- Context window. How much text the model can consider at once. Modern models handle 200K+ tokens (Claude Sonnet 4.6 handles 1M tokens).
- Tool use. Modern models can call functions you define, enabling agent behavior.
- Reasoning depth. Larger / more capable models reason better at complex problems but cost more per call.
- Latency. Smaller models respond in milliseconds; larger models can take seconds.
Example
In a customer support workflow, the LLM reads an inbound ticket, classifies it by topic + urgency, retrieves relevant docs from a knowledge base, drafts a response grounded in those docs, and either sends it or escalates to a human depending on confidence. The LLM is doing the thinking; the rest of the system is doing the orchestration.
Apply this
Get a real automation audit on your business.
Pay $4,500, fill an intake, get a Claude Opus-written strategic roadmap inside a week.
Start the audit →