GroqTools allows an Agent to interact with the Groq API for performing fast audio transcription, translation, and text-to-speech (TTS).

​
Prerequisites

Before using GroqTools, ensure you have the groq library installed and your Groq API key configured.

  1. Install the library:

    pip install -U groq

  2. Set your API key: Obtain your API key from the Groq Console and set it as an environment variable.

    export GROQ_API_KEY="your-groq-api-key"

​
Initialization

Import GroqTools and add it to your Agent’s tool list.

from agno.agent import Agent
from agno.tools.models.groq import GroqTools

agent = Agent(
    instructions=[
        "You are a helpful assistant that can transcribe audio, translate text and generate speech."
    ],
    tools=[GroqTools()],
    show_tool_calls=True,
)

​
Usage Examples

​
1. Transcribing Audio

This example demonstrates how to transcribe an audio file hosted at a URL.

transcription_agent.py
import os
from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.models.groq import GroqTools

audio_url = "https://agno-public.s3.amazonaws.com/demo_data/sample_conversation.wav"

agent = Agent(
    name="Groq Transcription Agent",
    model=OpenAIChat(id="gpt-4o"),
    tools=[GroqTools()],
    show_tool_calls=True,
)

agent.print_response(f"Please transcribe the audio file located at '{audio_url}'")

​
2. Translating Audio and Generating Speech

This example shows how to translate an audio file (e.g., French) to English and then generate a new audio file from the translated text.

translation_agent.py
from pathlib import Path
from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.models.groq import GroqTools
from agno.utils.media import save_base64_data

local_audio_path = "tmp/sample-fr.mp3"
output_path = Path("tmp/sample-en.mp3")
output_path.parent.mkdir(parents=True, exist_ok=True)

agent = Agent(
    name="Groq Translation Agent",
    model=OpenAIChat(id="gpt-4o-mini"),
    tools=[GroqTools()],
    show_tool_calls=True,
)

instruction = (
    f"Translate the audio file at '{local_audio_path}' to English. "
    f"Then, generate a new audio file using the translated English text."
)
agent.print_response(instruction)

response = agent.run_response
if response and response.audio:
    save_base64_data(response.audio[0].base64_audio, output_path)

You can customize the underlying Groq models used for transcription, translation, and TTS during initialization:

groq_tools = GroqTools(
    transcription_model="whisper-large-v3",
    translation_model="whisper-large-v3",
    tts_model="playai-tts",
    tts_voice="Chip-PlayAI"
)

​
Toolkit Functions

The GroqTools toolkit provides the following functions:

FunctionDescription
transcribe_audioTranscribes audio from a local file path or a public URL using Groq Whisper.
translate_audioTranslates audio from a local file path or public URL to English using Groq.
generate_speechGenerates speech from text using Groq TTS.

​
Developer Resources