Fireworks AI

Overview

The FireworksLLMService integrates Fireworks AI's lightning-fast inference for open models. It is particularly well-suited for high-throughput applications.

Installation

pip install piopiy-ai

Prerequisites

• A Fireworks AI API key (Get yours here).
• Set your API key in your environment:

  export FIREWORKS_API_KEY="your_api_key_here"

Configuration

FireworksLLMService Parameters

Parameter	Type	Default	Description
api_key	str	Required	Your Fireworks API key.
model	str	"accounts/fireworks/models/firefunction-v2"	Model identifier.
base_url	str	"https://api.fireworks.ai/inference/v1"	API endpoint.

Usage

Basic Setup

import os
from piopiy.services.fireworks.llm import FireworksLLMService

llm = FireworksLLMService(
    api_key=os.getenv("FIREWORKS_API_KEY"),
    model="accounts/fireworks/models/llama-v3p1-70b-instruct"
)

Notes

• Speed: Fireworks AI is optimized for ultra-low latency, making it a top choice for real-time AI agents.
• Function Calling: The firefunction-v2 model is specifically tuned for reliable tool use.