# LLM

> Bind model descriptors to provider clients while keeping credentials outside model and request data.

- Path: `Modules > LLM - Provider Clients`
- Human: https://looprig.com/docs/modules/llm
- Machine index: https://looprig.com/llms.txt

## Repository

| Field | Value |
| --- | --- |
| Repository | `github.com/looprig/llm` |
| Version | `v0.14.0` |
| GitHub | [looprig/llm](https://github.com/looprig/llm) |

## Description

Bind model descriptors to provider clients while keeping credentials outside model and request data.

## Where it fits

LLM is useful on its own when an application wants ready provider clients backed by explicit model descriptors and credentials. Within Looprig, it turns [Inference](/docs/modules/inference.md) contracts and [Credentials](/docs/modules/credentials.md) into clients that [Harness](/docs/modules/harness.md) applications can select. LLM owns provider construction; Inference retains the neutral request and response model.

## Dependencies

- [Core](/docs/modules/core.md)
- [Credentials](/docs/modules/credentials.md)
- [Inference](/docs/modules/inference.md)
- [Secrets](/docs/modules/secrets.md)

## Dependents

None.