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

# Quickstart

> Get from zero to your first OverControl API response in a few minutes.

In this guide, you will create an API key, configure your environment, choose a model, and send your first request.

<Note>
  Already using an OpenAI-compatible client or developer tool? OverControl supports compatible provider configurations. See [Integrations](/integrations) after completing this quickstart.
</Note>

***

## Before you start

You need:

* an [OverControl account](https://app.overcontrolgroup.com/register);
* an [OverControl API key](https://app.overcontrolgroup.com/dashboard);
* API credit available for your requests;
* a terminal with curl.

<Note>
  Every new OverControl API account includes free quota, so you can make your first requests without adding credit.
</Note>

<Card title="Open the API dashboard" icon="key" href="https://app.overcontrolgroup.com/dashboard" cta="Open dashboard">
  Create an API key, manage your API access, and review usage from the OverControl dashboard.
</Card>

***

## Step 1: Create an API key

Open the [OverControl dashboard](https://app.overcontrolgroup.com/dashboard) and create a new API key.

Copy the key when it is created and store it somewhere secure. Your application will use this key to authenticate requests to OverControl.

<Warning>
  Treat your API key like a password. Never commit it to source control, expose it in client-side code, or share it in screenshots, public chats, or support messages.
</Warning>

***

## Step 2: Configure your environment

Store your API key in an environment variable instead of placing it directly in your application code.

```bash theme={null}
export OVERCONTROL_API_KEY="your_api_key_here"
```

Set the OverControl Base URL used by your API account:

```bash theme={null}
export OVERCONTROL_BASE_URL="<OVERCONTROL_BASE_URL>"
```

Your application can now read these values without hard-coding credentials into the source code.

<Tip>
  Use a secrets manager or your deployment platform's encrypted environment variables when moving from local development to production.
</Tip>

***

## Step 3: Choose a model

For your first request, use:

```text theme={null}
lume-3
```

lume-3 is a strong starting point for general text generation, coding, analysis, and application workloads.

Other models are available for workloads that need different capabilities or performance characteristics.

***

## Step 4: Make your first request

OverControl supports an OpenAI-compatible API configuration. Send a request using your API key, Base URL, and model ID.

```bash theme={null}
curl "$OVERCONTROL_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $OVERCONTROL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "lume-3",
    "messages": [
      {
        "role": "user",
        "content": "Explain what an API is in one sentence."
      }
    ]
  }'
```

A successful request returns a model response in JSON.

The generated text is available in the assistant message returned by the API.

<Note>
  OverControl processes API content through European infrastructure. Prompts and responses are not retained as conversation history or reused for model training. See [Privacy and Security](/privacy-security) for details about the request lifecycle.
</Note>

***

## Step 5: Try another prompt

Once the first request works, change the message content and send another request.

For example:

```json theme={null}
{
  "role": "user",
  "content": "Give me three names for a developer tool that monitors API latency."
}
```

You can use the same request pattern for tasks such as:

* text generation;
* summarization;
* information extraction;
* coding assistance;
* classification;
* document processing workflows;
* reasoning and analysis.

Your application controls the context sent with each request.

***

## Use another model

Change the model value to route a request to another OverControl model.

For example:

```json theme={null}
{
  "model": "lume-3-max"
}
```

Use the Models guide to compare models before selecting one for a production workload.

<Tip>
  Choose models per workload rather than using the largest model for every request. This lets you balance capability, latency, and cost without changing the rest of your integration.
</Tip>

***

## Use an OpenAI-compatible client

If your application already uses an OpenAI-compatible provider, you can connect it to OverControl by changing the provider configuration.

You typically need four values:

| Setting  | Value                                  |
| -------- | -------------------------------------- |
| Provider | OpenAI compatible                      |
| Base URL | Your OverControl Base URL              |
| API key  | Your OverControl API key               |
| Model    | An OverControl model ID such as lume-3 |

This makes it possible to bring OverControl into existing applications and supported developer tools without rebuilding the surrounding workflow.

Explore [Integrations](/integrations)

***

## Before going to production

Your first request proves that the connection works. Before using OverControl in a production application:

* keep API keys on the server side;
* store secrets outside your source code;
* choose the appropriate model for the workload;
* review current model pricing;
* define application-level timeouts and error handling;
* review how sensitive data is processed;
* monitor API usage and spending.

<CardGroup cols={2}>
  <Card title="Models" icon="brain" href="/models">
    Compare models and choose the right one for your workload.
  </Card>

  <Card title="Privacy and security" icon="shield" href="/privacy-security">
    Understand how API requests and sensitive data are handled.
  </Card>

  <Card title="Integrations" icon="plug" href="/integrations">
    Connect OverControl to supported clients and developer tools.
  </Card>

  <Card title="Pricing" icon="credit-card" href="https://overcontrolgroup.com/pricing">
    Review current input, cached input, and output token pricing.
  </Card>
</CardGroup>

***

## Next steps

You have now completed the basic OverControl API flow:

<Steps>
  <Step title="Choose the right model">
    Compare model capabilities, context windows, and pricing for your workload.

    [Explore models](/models)
  </Step>

  <Step title="Understand privacy and security">
    Review how requests are processed, protected, and removed after processing.

    [Read privacy and security](/privacy-security)
  </Step>

  <Step title="Connect your existing tools">
    Configure supported developer tools and OpenAI-compatible clients to use OverControl.

    [Explore integrations](/integrations)
  </Step>
</Steps>