Skip to main content
Access ultra-fast inference powered by Groq’s LPU™ (Language Processing Unit) Inference Engine for various open-source models.

​
Configuration

Configuration examples for ~/.factory/config.json: 
{
  "custom_models": [
    {
      "model_display_name": "Kimi K2 [Groq]",
      "model": "moonshotai/kimi-k2-instruct-0905",
      "base_url": "https://api.groq.com/openai/v1",
      "api_key": "YOUR_GROQ_KEY",
      "provider": "generic-chat-completion-api",
      "max_tokens": 32000
    }
  ]
}

​
Getting Started

  1. Sign up at groq.com
  2. Get your API key from console.groq.com
  3. Browse available models in the Groq documentation
  4. Add desired models to your configuration

​
Notes

  • Base URL format: https://api.groq.com/openai/v1
  • Groq uses the generic-chat-completion-api provider type
  • Known for extremely fast inference speeds thanks to LPU architecture
  • Supports streaming responses for compatible models