> ## Documentation Index
> Fetch the complete documentation index at: https://docs.lstack.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Providers and models

> Configure the model endpoints and capacity used by LStack workers.

A provider tells LStack how workers reach a model service. It contains the endpoint, authentication mode, model configuration, capability information, and concurrency limits used when workers are created and scheduled.

## Provider responsibilities

* Identify the model endpoint and authentication method
* Define the model used by workers
* Limit concurrent model work
* Validate connectivity and relevant capabilities
* Provide credentials to authorized workers at assignment time

## Current execution compatibility

Current qualified workers run Claude Code and require a provider that supports the Anthropic Messages API shape. Local or self-hosted compatible endpoints and supported cloud configurations can be represented as provider profiles.

<Info>
  A saved provider is not automatically compatible with every worker engine. Compatibility depends on the engine and API surface supported by the provider.
</Info>

## Capacity

Provider concurrency limits and worker policy work together. Creating additional workers does not increase the capacity of the underlying model endpoint.

<Card title="Workers" icon="bot" href="/workers/overview">
  Create and operate execution units backed by a configured provider.
</Card>