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

# Managed Models

> Choose which AI you want to power your tasks

<Card title=" " img="https://mintcdn.com/aithing/ymZuY2hcYq_i4O0e/images/v1/managed-models.png?fit=max&auto=format&n=ymZuY2hcYq_i4O0e&q=85&s=d7aa8f84c38de4f205f341d2a00ae66a" width="1048" height="864" data-path="images/v1/managed-models.png">
  Connect to different large language models (brains) with your own API key—giving you full control over your AI capabilities.
</Card>

## Managed Models `BYOK`

Managed models allow you to select AI models based on your needs using your **API key**, such as:

* **Fast and cost-effective** (Claude Haiku 3.5)
* **Balanced** (Claude Sonnet 4)
* **High intelligence** (Claude Opus 4.1)

To connect:

1. Get an API key from [Anthropic](https://docs.anthropic.com/claude/reference/getting-started-with-the-api)
2. Paste the key into the field that starts with `sk-ant-...`

This option gives you **full ownership and flexibility**, with no [AI Thing](https://aithing.dev) plan required.

## Model Settings

In **Settings → Preferences**, you can fine-tune how models behave:

<Frame>
  <img src="https://mintcdn.com/aithing/ymZuY2hcYq_i4O0e/images/v1/preferences-models.png?fit=max&auto=format&n=ymZuY2hcYq_i4O0e&q=85&s=865c2276e8a74a4a261ede357c436f2c" width="1052" height="354" data-path="images/v1/preferences-models.png" />
</Frame>

### Prompt Cache

Reuses common parts of a prompt to speed up responses and lower costs during long or repetitive conversations.
For one-time or short queries, caching may actually be more expensive — it’s most useful when a conversation is expected to continue over many turns.
[Read more](https://docs.claude.com/en/docs/build-with-claude/prompt-caching).

### Output Tokens

Set the maximum length of responses. Useful for keeping outputs concise or allowing longer, more detailed answers.