Large Language Model (LLM)
An LLM (large language model) is a generative AI model that learns statistical patterns from massive text corpora, allowing it to write, answer, and reason in language within its context window.
Technically, most are built on transformer architecture and self-attention. Limitations include hallucination, AI bias, and lack of knowledge about events after training cutoff. Explore ChatGPT, Claude, Gemini, and OpenRouter for model options.
Key characteristics
- Is a model class trained on large text datasets to understand and generate natural language.
- Can solve many task types but is sensitive to prompt quality, context, and hallucinations.
- Is a core component in modern AI products, but often needs tools and external data.
Example
An LLM can draft a report in seconds,
but the output becomes more reliable when the model gets internal data, clear instructions, and citation requirements.