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

# Groq

> Use ultra-fast LLM inference with Groq

## Overview

Groq provides blazing-fast inference for popular open-source models through their custom LPU hardware. Known for exceptional speed while maintaining strong model quality.

## Available Models

<CardGroup cols={2}>
  <Card title="DeepSeek Llama (70B)" icon="brain">
    **2 credits** • Distilled reasoning model

    * 64K context window
    * Excellent logical capabilities
    * Speed: Fast • Cost: Low
    * **Reasoning model** optimized for efficiency
  </Card>

  <Card title="Llama 3.3 Versatile (70B)" icon="sparkles">
    **3 credits** • Versatile large model

    * 128K context window
    * Excellent reasoning across diverse tasks
    * Speed: Fast • Cost: Medium
    * Best for: General-purpose applications
  </Card>

  <Card title="Llama 3.1 Instant (8B)" icon="zap">
    **1 credit** • Ultra-fast small model

    * 128K context window
    * Good reasoning with minimal latency
    * Speed: Extremely Fast • Cost: Very Low
    * Best for: Real-time applications
  </Card>

  <Card title="Qwen QWQ (32B)" icon="brain">
    **2 credits** • Multilingual reasoning

    * 32K context window
    * Excellent reasoning and multilingual
    * Speed: Fast • Cost: Low
    * **Reasoning model** for diverse tasks
  </Card>

  <Card title="Qwen 3 (32B)" icon="bolt">
    **2 credits** • Next-gen instruction model

    * 32K context window
    * Excellent instruction-following
    * Speed: Fast • Cost: Low
    * Best for: Strong language tasks
  </Card>

  <Card title="Mistral Saba (24B)" icon="circle-check">
    **2 credits** • Reasoning and chat

    * 32K context window
    * Excellent reasoning
    * Speed: Fast • Cost: Low
    * Best for: Balanced performance
  </Card>

  <Card title="LLaMA 4 Maverick (17B 128E)" icon="crown">
    **3 credits** • 128 experts model

    * 128K context window
    * Excellent for advanced instructions
    * Speed: Medium • Cost: Medium
    * Best for: Complex instruction tasks
  </Card>

  <Card title="LLaMA 4 Scout (17B 16E)" icon="search">
    **1 credit** • 16 experts model

    * 128K context window
    * Good for efficient inference
    * Speed: Fast • Cost: Low
    * Best for: Lighter workloads
  </Card>
</CardGroup>

## Setup

### Using BoostGPT-Hosted API Keys

<Steps>
  <Step title="Select Groq Model">
    In your BoostGPT dashboard, select any Groq model when creating or configuring your bot.
  </Step>

  <Step title="Choose Your Model">
    * **Llama 3.1 Instant**: For real-time, ultra-fast responses
    * **Llama 3.3 Versatile**: For balanced performance
    * **DeepSeek Llama 70B**: For reasoning tasks
    * **Qwen models**: For multilingual applications
  </Step>
</Steps>

### Using Your Own Groq API Key

<Tabs>
  <Tab title="Dashboard Setup">
    <Steps>
      <Step title="Navigate to Integrations">
        Go to [app.boostgpt.co](https://app.boostgpt.co/integrations) and select **Integrations**
      </Step>

      <Step title="Select Groq">
        Find and click on the **Groq** provider
      </Step>

      <Step title="Add API Key">
        Enter your Groq API key and select which agents will use this key
      </Step>

      <Step title="Save Configuration">
        Click save to apply your custom API key
      </Step>
    </Steps>
  </Tab>

  <Tab title="Core SDK">
    ```javascript bot.js theme={null}
    import { BoostGPT } from 'boostgpt';

    const client = new BoostGPT({
      project_id: process.env.BOOSTGPT_PROJECT_ID,
      key: process.env.BOOSTGPT_API_KEY
    });

    // Create bot with Groq model
    const botResponse = await client.createBot({
      name: 'My Groq Bot',
      model: 'llama-3.1-8b-instant', // Ultra-fast!
      instruction: 'You are a quick and helpful assistant.',
      max_reply_tokens: 1000,
      status: 'active'
    });

    if (botResponse.err) {
      console.error('Error:', botResponse.err);
    } else {
      console.log('Bot created:', botResponse.response);
    }
    ```
  </Tab>
</Tabs>

## Performance Benefits

<Tip>
  Groq's LPU (Language Processing Unit) hardware provides inference speeds up to 10x faster than traditional GPUs, making it ideal for real-time applications.
</Tip>

## Next Steps

<CardGroup cols={2}>
  <Card title="Reasoning Models" icon="brain" href="/providers/reasoning-models">
    Learn about reasoning models on Groq
  </Card>

  <Card title="Model Comparison" icon="scale" href="/providers/model-comparison">
    Compare Groq with other providers
  </Card>

  <Card title="SDK Reference" icon="code" href="/sdk/core/api-reference">
    Full API documentation
  </Card>

  <Card title="Bring Your Own Key" icon="key" href="/providers/bring-your-own-keys">
    Use your own Groq API key
  </Card>
</CardGroup>