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

# Pezzo Client - Python

The Pezzo client is a PyPi that allows you to easily integrate your application with Pezzo.

## Getting Started

### Intall the Pezzo Client

Install the [pezzo](https://pypi.org/project/pezzo/) package from PyPi:

<CodeGroup>
  ```bash pip theme={null}
  pip install pezzo
  ```

  ```bash poetry theme={null}
  poetry add pezzo
  ```
</CodeGroup>

### Initialize the Pezzo Client

To initialize the Pezzo client, simply import it into your script. It will automatically read the configuration from your environment variables:

* `PEZZO_API_KEY`: Your Pezzo API key
* `PEZZO_PROJECT_ID`: Your Pezzo project ID
* `PEZZO_ENVIRONMENT`: The environment you want to use (e.g. `Production`, which is the default environment created by Pezzo)

<CodeGroup>
  ```py main.py theme={null}
  from pezzo.client import pezzo
  ```
</CodeGroup>

In the above example, we created a `libs/pezzo.ts` file in which we instantiate the Pezzo client and export it. We can then import it in other areas of our application.

<CardGroup cols={2}>
  <Card title="Use Pezzo with OpenAI" icon="bolt-lightning" href="/client/integrations/openai">
    Learn how to use Pezzo to observe and manage your OpenAI API calls.
  </Card>
</CardGroup>

## API Reference

<ResponseField name="pezzo.get_prompt(prompt_name: string)" type="Function">
  <div style={{ marginLeft: 20 }}>
    <ParamField path="prompt_name" type="string">
      The name of the prompt to retrieve. The prompt must be deployed to the current environment specified when initializing the Pezzo client.
    </ParamField>
  </div>
</ResponseField>