A thoughtful AI collaborator for writing, analysis, coding, and long-form work.
Full evaluationGlossary · Updated Sep 2, 2026
Large language model (LLM)
A neural network trained on vast text corpora that predicts and generates language, powering chat assistants and writing tools.
Definition
A large language model is a transformer-based neural network trained on enormous amounts of text to predict the next token. Through that objective it learns grammar, facts, reasoning patterns, and styles, which lets it answer questions, draft documents, write code, and follow instructions. Products such as ChatGPT, Claude, and Gemini are interfaces on top of LLMs, often combined with tools, retrieval, and safety layers.
Why it matters when choosing a tool
Almost every AI tool in this directory is built on an LLM, so understanding their strengths, such as fluent generation, and their failure modes, such as hallucination, helps you judge what a product can reliably do.
Where you will meet it
AI Assistants, AI Writing & Editing, AI Coding & Development
Related terms
Tools where this matters
Reviewed products in the related categories
A broad multimodal assistant for research, creation, analysis, and agentic work.
Full evaluation
An answer engine built around current web research and visible citations.
Full evaluation
AI writing, search, and task assistance inside a connected workspace and knowledge base.
Full evaluation
An AI-native code editor for repository-aware agents, edits, review, and automation.
Full evaluation
Anthropic's agentic coding tool that works in the terminal, IDE, desktop app, and browser to plan and execute multi-step changes.
Full evaluation