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

# Groq

> Leverage Groq's LPU inference engine for blisteringly fast responses.

<Frame>
  <img src="https://mintcdn.com/arivu/piekGGEpY74vFGfF/images/groq.png?fit=max&auto=format&n=piekGGEpY74vFGfF&q=85&s=732256c253711874fae528102ecc778f" alt="Groq" width="1919" height="912" data-path="images/groq.png" />
</Frame>

## Overview

Groq operates using an LPU™ (Language Processing Unit) architecture, specifically engineered to deliver real-time AI capabilities with latency far below traditional GPU deployments.

<CardGroup cols={2}>
  <Card title="Unmatched Speed" icon="rocket">
    Hundreds of tokens per second for popular models like Llama 3.
  </Card>

  <Card title="Open Source Models" icon="code-branch">
    First-class support for Llama, Mixtral, and Gemma model variants.
  </Card>
</CardGroup>

## Configuration

<Steps>
  <Step title="Get your API Key">
    Sign up on the [GroqCloud Console](https://console.groq.com/).
  </Step>

  <Step title="Environment Settings">
    Add the credentials to your environment:

    <CodeGroup>
      ```bash .env theme={null}
      GROQ_API_KEY="gsk_..."
      DEFAULT_MODEL="llama3-70b-8192"
      ```
    </CodeGroup>
  </Step>
</Steps>

<Info>Because of Groq's incredible generation speed, it is highly recommended for streaming applications (e.g. voice or real-time chatbots).</Info>