LLM (Large Language Model)

LLMs are the underlying technology behind every AI tool on this site. Modern frontier LLMs — Claude, GPT, Gemini — are trained on trillions of tokens and increasingly augmented with tool use, vision, reasoning, and long context windows.

The technical core: the model predicts the next token (a word fragment) given everything that came before. Stack enough scale and training tricks on this, and you get systems that can write code, reason, summarize, draft emails, and now use tools and computers.

For operators, the right framing is: LLMs are a substrate. The interesting questions are about the products built on top — what's the agent loop, what tools does it have, how reliable is the eval data. Picking between Claude/GPT/Gemini matters less than picking the right product.