# Large language model (LLM) | IonBin

> A large language model is an AI system trained on very large amounts of text to predict the next piece of text.

Canonical page: https://ionbin.com/wiki/concepts/large-language-model/
Agent guide: https://ionbin.com/llms.txt

This is a readable version of the published page. Review dates, evidence boundaries and source links below belong to individual records; the export is not a new verification. Examples are reference material, not instructions to execute.

Concept

## Large language model (LLM)

A large language model is an AI system trained on very large amounts of text to predict the next piece of text. That simple skill lets it draft, summarize, translate, classify and work through problems. [Claude](https://ionbin.com/tools/claude/), GPT, Gemini and [Grok](https://ionbin.com/tools/grok/) are examples of LLM families. [ChatGPT](https://ionbin.com/tools/chatgpt/) and similar apps are products built on top of them.

Verified Oct 18, 2026 Review by Apr 16, 2027 3 sources

### What it is

An LLM is a neural network trained on large collections of text, and often images and audio too. It works in [tokens](https://ionbin.com/wiki/concepts/tokens/), small chunks of text. For text generation, it predicts successive tokens using the material supplied in its [context window](https://ionbin.com/wiki/concepts/context-window/). IBM groups LLMs under [generative AI](https://ionbin.com/wiki/concepts/generative-ai/): models that create new content in response to a prompt.

### How it works

Two phases matter for business users:

| Phase | What happens | Who does it | How often |

| — | — | — | — |

| Training | Pretraining learns patterns; later training can adapt behavior or skills. A published [knowledge cutoff](https://ionbin.com/wiki/concepts/knowledge-cutoff/) describes training-data recency, not guaranteed knowledge of every earlier fact. | The model developer or a team adapting it | Multiple stages; releases and updates vary |

| Use ([inference](https://ionbin.com/wiki/concepts/inference/)) | You send a prompt and the model predicts a response. Apps add memory, search and tools around it. | You, every request | Every time |

Vendors publish model lists with context sizes, cutoffs and prices. Anthropic and OpenAI both document theirs.

### Benefits

- General-purpose: one model handles writing, analysis, coding and more.

- Works in plain language, so non-technical staff can use it.

- Can be connected to your data and tools to do real work.

### Limits and risks

- Can state false things confidently (see [hallucination](https://ionbin.com/wiki/concepts/hallucination/)).

- A cutoff is not live access or a completeness guarantee. Supply current evidence for changing facts; search and documents can also contain errors.

- Output varies between runs and between models.

### Agent compatibility

Not tool-specific. See [Agent-ready](https://ionbin.com/agent-ready/) for connector support by tool.

Levels come from each tool's recorded, sourced compatibility. [How the levels work](https://ionbin.com/wiki/questions/connect-a-tool-to-an-ai-agent/).

### Related entries

[Tokens · Concept](https://ionbin.com/wiki/concepts/tokens/)[Context window · Concept](https://ionbin.com/wiki/concepts/context-window/)[Generative AI · Concept](https://ionbin.com/wiki/concepts/generative-ai/)[Which AI model should I use for a business task? · Question](https://ionbin.com/wiki/questions/which-ai-model-should-i-use/)[Claude (Anthropic) · Model](https://ionbin.com/wiki/models/claude/)[Knowledge cutoff · Concept](https://ionbin.com/wiki/concepts/knowledge-cutoff/)[Multimodal AI · Concept](https://ionbin.com/wiki/concepts/multimodal-ai/)[Temperature · Concept](https://ionbin.com/wiki/concepts/temperature/)

### Sources

- [What is generative AI?](https://www.ibm.com/think/topics/generative-ai) · IBMreference accessed Oct 17, 2026

- [Models overview](https://docs.anthropic.com/en/docs/about-claude/models/overview) · Anthropicofficial accessed Oct 17, 2026

- [Models](https://platform.openai.com/docs/models) · OpenAIofficial accessed Oct 17, 2026

Practitioner views are restated in our words and labeled as views, not measured results. Product facts come from official sources dated above.

### Suggest an edit

IonBin's wiki isn't user-edited. Corrections go to our Directory Curator, who checks them against primary sources before anything changes. Spot something wrong or out of date on “Large language model (LLM)”? Email us with a link to a primary source.

[Suggest an edit by email](mailto:contact@ionbin.com?subject=Suggest%20an%20edit%3A%20Large%20language%20model%20(LLM)%20%5Blarge-language-model%5D&body=Page%3A%20https%3A%2F%2Fionbin.com%2Fwiki%2Fconcepts%2Flarge-language-model%2F%0AEntry%3A%20concepts%2Flarge-language-model%0A%0AWhat%20should%20change%3A%0A%0A%0APrimary%20source%20(link)%3A%0A) or write to [contact@ionbin.com](mailto:contact@ionbin.com)

Entry facts

**Type**: [Concepts & glossary](https://ionbin.com/wiki/concepts/)

**Status**: Reviewed

**Verified**: Oct 18, 2026

**Review by**: Apr 16, 2027 (every 180 days)

**Sources**: [2 official · 1 other](https://ionbin.com/wiki/concepts/large-language-model/#sources)

The wiki carries no sponsored placements or affiliate links. [Editorial policy](https://ionbin.com/editorial-policy/).
