Overview
GroqLLMService provides access to Groq’s language models through an OpenAI-compatible interface. It inherits from OpenAILLMService and supports streaming responses, function calling, and context management with ultra-fast inference speeds.
Groq LLM API Reference
Pipecat’s API methods for Groq integration
Example Implementation
Complete example with function calling
Groq Documentation
Official Groq API documentation and features
Groq Console
Access models and manage API keys
Installation
To use Groq services, install the required dependency:Prerequisites
Groq Account Setup
Before using Groq LLM services, you need:- Groq Account: Sign up at Groq Console
- API Key: Generate an API key from your console dashboard
- Model Selection: Choose from available models with ultra-fast inference
Required Environment Variables
GROQ_API_KEY: Your Groq API key for authentication
Configuration
str
required
Groq API key for authentication. Raises
ValueError when empty.str
default:"https://api.groq.com/openai/v1"
Base URL for Groq API endpoint.
str
default:"None"
deprecated
Deprecated since 0.0.105, removed in 2.0.0. Use
settings=GroqLLMService.Settings(model=...) instead.default_headers, service_tier,
retry_on_timeout, and retry_timeout_secs, as described for
OpenAILLMService.
Settings
Runtime-configurable settings passed via thesettings constructor argument using GroqLLMService.Settings(...). These can be updated mid-conversation with LLMUpdateSettingsFrame. See Service Settings for details.
This service extends OpenAILLMService settings with Groq-specific parameters. See OpenAI LLM Settings for inherited parameters.
The default model is "openai/gpt-oss-120b".
Literal["none", "default", "low", "medium", "high"] | None
default:"None"
Controls how much the model thinks before answering. Which values a model accepts varies:
"low", "medium" and "high" on the GPT-OSS models and qwen/qwen3.8-27b, "none" and "default" on the Qwen models, and the non-reasoning models reject the parameter altogether. "none" also keeps a Qwen model’s reasoning out of the spoken response, which Groq otherwise streams inline in <think> tags. When unset, Groq’s per-model default applies.Usage
Basic Setup
With Custom Settings
Notes
- Groq provides ultra-fast inference using custom LPU (Language Processing Unit) hardware.
- Groq fully supports the OpenAI-compatible parameter set inherited from
OpenAILLMService.